Project: Identify Customer Segments

In this project, you will apply unsupervised learning techniques to identify segments of the population that form the core customer base for a mail-order sales company in Germany. These segments can then be used to direct marketing campaigns towards audiences that will have the highest expected rate of returns. The data that you will use has been provided by our partners at Bertelsmann Arvato Analytics, and represents a real-life data science task.

This notebook will help you complete this task by providing a framework within which you will perform your analysis steps. In each step of the project, you will see some text describing the subtask that you will perform, followed by one or more code cells for you to complete your work. Feel free to add additional code and markdown cells as you go along so that you can explore everything in precise chunks. The code cells provided in the base template will outline only the major tasks, and will usually not be enough to cover all of the minor tasks that comprise it.

It should be noted that while there will be precise guidelines on how you should handle certain tasks in the project, there will also be places where an exact specification is not provided. There will be times in the project where you will need to make and justify your own decisions on how to treat the data. These are places where there may not be only one way to handle the data. In real-life tasks, there may be many valid ways to approach an analysis task. One of the most important things you can do is clearly document your approach so that other scientists can understand the decisions you've made.

At the end of most sections, there will be a Markdown cell labeled Discussion. In these cells, you will report your findings for the completed section, as well as document the decisions that you made in your approach to each subtask. Your project will be evaluated not just on the code used to complete the tasks outlined, but also your communication about your observations and conclusions at each stage.

# import libraries here; add more as necessary
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import plotly.express as px
import json
from sklearn.decomposition import PCA
from sklearn.cluster import KMeans
from sklearn.cluster import MiniBatchKMeans



# magic word for producing visualizations in notebook
%matplotlib inline

'''
Import note: The classroom currently uses sklearn version 0.19.
If you need to use an imputer, it is available in sklearn.preprocessing.Imputer,
instead of sklearn.impute as in newer versions of sklearn.
'''

'\nImport note: The classroom currently uses sklearn version 0.19.\nIf you need to use an imputer, it is available in sklearn.preprocessing.Imputer,\ninstead of sklearn.impute as in newer versions of sklearn.\n'

Step 0: Load the Data

There are four files associated with this project (not including this one):

• Udacity_AZDIAS_Subset.csv: Demographics data for the general population of Germany; 891211 persons (rows) x 85 features (columns).
• Udacity_CUSTOMERS_Subset.csv: Demographics data for customers of a mail-order company; 191652 persons (rows) x 85 features (columns).
• Data_Dictionary.md: Detailed information file about the features in the provided datasets.
• AZDIAS_Feature_Summary.csv: Summary of feature attributes for demographics data; 85 features (rows) x 4 columns

Each row of the demographics files represents a single person, but also includes information outside of individuals, including information about their household, building, and neighborhood. You will use this information to cluster the general population into groups with similar demographic properties. Then, you will see how the people in the customers dataset fit into those created clusters. The hope here is that certain clusters are over-represented in the customers data, as compared to the general population; those over-represented clusters will be assumed to be part of the core userbase. This information can then be used for further applications, such as targeting for a marketing campaign.

To start off with, load in the demographics data for the general population into a pandas DataFrame, and do the same for the feature attributes summary. Note for all of the .csv data files in this project: they're semicolon (;) delimited, so you'll need an additional argument in your read_csv() call to read in the data properly. Also, considering the size of the main dataset, it may take some time for it to load completely.

Once the dataset is loaded, it's recommended that you take a little bit of time just browsing the general structure of the dataset and feature summary file. You'll be getting deep into the innards of the cleaning in the first major step of the project, so gaining some general familiarity can help you get your bearings.

# Load in the general demographics data.
azdias = pd.read_csv('Udacity_AZDIAS_Subset.csv',sep=';')

# Load in the feature summary file.
feat_info = pd.read_csv('AZDIAS_Feature_Summary.csv',sep=';')

# Check the structure of the data after it's loaded (e.g. print the number of
# rows and columns, print the first few rows).
azdias.head()

	AGER_TYP	ALTERSKATEGORIE_GROB	ANREDE_KZ	CJT_GESAMTTYP	FINANZ_MINIMALIST	FINANZ_SPARER	FINANZ_VORSORGER	FINANZ_ANLEGER	FINANZ_UNAUFFAELLIGER	FINANZ_HAUSBAUER	...	PLZ8_ANTG1	PLZ8_ANTG2	PLZ8_ANTG3	PLZ8_ANTG4	PLZ8_BAUMAX	PLZ8_HHZ	PLZ8_GBZ	ARBEIT	ORTSGR_KLS9	RELAT_AB
0	-1	2	1	2.0	3	4	3	5	5	3	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
1	-1	1	2	5.0	1	5	2	5	4	5	...	2.0	3.0	2.0	1.0	1.0	5.0	4.0	3.0	5.0	4.0
2	-1	3	2	3.0	1	4	1	2	3	5	...	3.0	3.0	1.0	0.0	1.0	4.0	4.0	3.0	5.0	2.0
3	2	4	2	2.0	4	2	5	2	1	2	...	2.0	2.0	2.0	0.0	1.0	3.0	4.0	2.0	3.0	3.0
4	-1	3	1	5.0	4	3	4	1	3	2	...	2.0	4.0	2.0	1.0	2.0	3.0	3.0	4.0	6.0	5.0

5 rows × 85 columns

print(azdias.shape)

(891221, 85)

feat_info.head()

	attribute	information_level	type	missing_or_unknown
0	AGER_TYP	person	categorical	[-1,0]
1	ALTERSKATEGORIE_GROB	person	ordinal	[-1,0,9]
2	ANREDE_KZ	person	categorical	[-1,0]
3	CJT_GESAMTTYP	person	categorical	[0]
4	FINANZ_MINIMALIST	person	ordinal	[-1]

print(feat_info.shape)

(85, 4)

I can see that dataset AZDIAS_Feature_Summary presents basic information about data in columns of dataset Udacity_AZDIAS_Subset.

Tip: Add additional cells to keep everything in reasonably-sized chunks! Keyboard shortcut esc --> a (press escape to enter command mode, then press the 'A' key) adds a new cell before the active cell, and esc --> b adds a new cell after the active cell. If you need to convert an active cell to a markdown cell, use esc --> m and to convert to a code cell, use esc --> y.

Step 1: Preprocessing

Step 1.1: Assess Missing Data

The feature summary file contains a summary of properties for each demographics data column. You will use this file to help you make cleaning decisions during this stage of the project. First of all, you should assess the demographics data in terms of missing data. Pay attention to the following points as you perform your analysis, and take notes on what you observe. Make sure that you fill in the Discussion cell with your findings and decisions at the end of each step that has one!

Step 1.1.1: Convert Missing Value Codes to NaNs

The fourth column of the feature attributes summary (loaded in above as feat_info) documents the codes from the data dictionary that indicate missing or unknown data. While the file encodes this as a list (e.g. [-1,0]), this will get read in as a string object. You'll need to do a little bit of parsing to make use of it to identify and clean the data. Convert data that matches a 'missing' or 'unknown' value code into a numpy NaN value. You might want to see how much data takes on a 'missing' or 'unknown' code, and how much data is naturally missing, as a point of interest.

As one more reminder, you are encouraged to add additional cells to break up your analysis into manageable chunks.

FIRSTLY, BEFORE I take care for the missing values, I would like to see (and present in the notebook) columns in two categories:

a) categorical and ordinal columns - and for each of them to show all unique values in the form of **pie chart** b) continuous-valued columns - and for each of them print **histogram**. **I know that information about the type of data in the columns and the individual values that represent the missing values for each column is in the Summary file, but I want to make sure and do this analysis by myself.**

# Let's take a help from provided info file. How this file describe main dataset's columns?
print(feat_info['type'].unique())

['categorical' 'ordinal' 'numeric' 'mixed' 'interval']

I'm not strongly sure what's behind these criteria, so I'll take a sample for previewing...

# Looking for columns which are 'categorical'
print(feat_info[feat_info['type'] == 'categorical'][:3], '\n')

# Looking for columns which are 'ordinal'
print(feat_info[feat_info['type'] == 'ordinal'][:3], '\n')

# Looking for columns which are 'numeric'
print(feat_info[feat_info['type'] == 'numeric'][:3], '\n')

# Looking for columns which are 'mixed'
print(feat_info[feat_info['type'] == 'mixed'][:3], '\n')

# Looking for columns which are 'interval'
print(feat_info[feat_info['type'] == 'interval'][:3], '\n')

       attribute information_level         type missing_or_unknown
0       AGER_TYP            person  categorical             [-1,0]
2      ANREDE_KZ            person  categorical             [-1,0]
3  CJT_GESAMTTYP            person  categorical                [0] 

              attribute information_level     type missing_or_unknown
1  ALTERSKATEGORIE_GROB            person  ordinal           [-1,0,9]
4     FINANZ_MINIMALIST            person  ordinal               [-1]
5         FINANZ_SPARER            person  ordinal               [-1] 

       attribute information_level     type missing_or_unknown
11   GEBURTSJAHR            person  numeric                [0]
44  ANZ_PERSONEN         household  numeric                 []
45     ANZ_TITEL         household  numeric                 [] 

                attribute information_level   type missing_or_unknown
15    LP_LEBENSPHASE_FEIN            person  mixed                [0]
16    LP_LEBENSPHASE_GROB            person  mixed                [0]
22  PRAEGENDE_JUGENDJAHRE            person  mixed             [-1,0] 

   attribute information_level      type missing_or_unknown
43  ALTER_HH         household  interval                [0] 

For finded examples of columns, I'll print some values that are inside of them, but also, outside of this notebook, I'll inspect "Data_Dictionary.md" file for detailed information and better understanding.

# I have samples of column names for each data category, now let's see how the data looks like in those groups of columns

# data in columns which are 'categorical'
print(azdias['AGER_TYP'][:5], '\n')
print(azdias['ANREDE_KZ'][:5], '\n')
print(azdias['CJT_GESAMTTYP'][:5], '\n')

0   -1
1   -1
2   -1
3    2
4   -1
Name: AGER_TYP, dtype: int64 

0    1
1    2
2    2
3    2
4    1
Name: ANREDE_KZ, dtype: int64 

0    2.0
1    5.0
2    3.0
3    2.0
4    5.0
Name: CJT_GESAMTTYP, dtype: float64 

Here I was fairly sure, because the name "categorical" makes it clear how we should treat this data. In the above-printed samples, I see digital labels in int and float formats. I was expecting string formats but you can't judge that there are no such columns either. The "categorical" columns will be converted to one-hot-encoder.

# data in columns which are 'ordinal'
print(azdias['ALTERSKATEGORIE_GROB'][:5], '\n')
print(azdias['FINANZ_MINIMALIST'][:5], '\n')
print(azdias['FINANZ_SPARER'][:5], '\n')

0    2
1    1
2    3
3    4
4    3
Name: ALTERSKATEGORIE_GROB, dtype: int64 

0    3
1    1
2    1
3    4
4    4
Name: FINANZ_MINIMALIST, dtype: int64 

0    4
1    5
2    4
3    2
4    3
Name: FINANZ_SPARER, dtype: int64 

The name "ordinal" also clearly identifies the data in the columns. These are categorical data, but such that there are dependencies and ranks between different values. Therefore, for such data, I will keep the original value numbering (at least at the stage of data cleansing, before scaling).

# data in columns which are 'numeric'
print(azdias['GEBURTSJAHR'][:5], '\n')
print(azdias['ANZ_PERSONEN'][:5], '\n')
print(azdias['ANZ_TITEL'][:5], '\n')

0       0
1    1996
2    1979
3    1957
4    1963
Name: GEBURTSJAHR, dtype: int64 

0    NaN
1    2.0
2    1.0
3    0.0
4    4.0
Name: ANZ_PERSONEN, dtype: float64 

0    NaN
1    0.0
2    0.0
3    0.0
4    0.0
Name: ANZ_TITEL, dtype: float64 

"Numeric" columns are a combination of continuous values and discrete but strictly numerical values such as year (1974, 1999 etc.). In order not to complicate the work too much, I will not divide this set into actual continuous and ordinal data, but will treat all of them as if they were continuous data. It follows from this decision that when I further inspect the contents of the columns, I will not print or graph the values unique to those columns. To find missing or incorrect values, I will only look for anything that is not a number or for "-1" if it is so indicated by the sources.

# data in columns which are 'mixed'
print(azdias['LP_LEBENSPHASE_FEIN'][:5], '\n')
print(azdias['LP_LEBENSPHASE_GROB'][:5], '\n')
print(azdias['PRAEGENDE_JUGENDJAHRE'][:5], '\n')

0    15.0
1    21.0
2     3.0
3     0.0
4    32.0
Name: LP_LEBENSPHASE_FEIN, dtype: float64 

0     4.0
1     6.0
2     1.0
3     0.0
4    10.0
Name: LP_LEBENSPHASE_GROB, dtype: float64 

0     0
1    14
2    15
3     8
4     8
Name: PRAEGENDE_JUGENDJAHRE, dtype: int64 

I'll rather treat data in columns "mixed" as categorical data, but confirm that after visual inspection of values in those columns.

# data in columns which are 'interval'
print(azdias['ALTER_HH'][:5], '\n')

0     NaN
1     0.0
2    17.0
3    13.0
4    20.0
Name: ALTER_HH, dtype: float64 

There is only one column with 'interval' data, and I've learned from "Data_Dictionary" file, that these values are discreet numeric values for birthdate intervals. I'll treat that column as "ordinal" data.

So, as I wrote above, I would like to split the dataset columns into two groups:

a) "categorical" and "ordinal" columns

b) and columns with continuous numeric values

Based on the review above and in line with the categories from AZDIAS_Feature_Summary.csv:

a) == 'categorical', 'ordinal', 'mixed', 'interval'

b) == 'numeric'

# "Visual" identify of missing / unknown / aberrant values (and all values in general... :) )
# I only display columns from the "a)" group in the form of a pie chart


# making list of names of "a)" group columns
list_of_categorical_cols = feat_info[feat_info['type']!='numeric']['attribute']
print(list_of_categorical_cols, '\n')


for col_name in list_of_categorical_cols: 
    fig = px.pie(azdias[col_name].value_counts(dropna=False), values=col_name, 
             names = azdias[col_name].value_counts(dropna=False).index, title = col_name, template='ggplot2') 
    
    fig.show()

0                 AGER_TYP
1     ALTERSKATEGORIE_GROB
2                ANREDE_KZ
3            CJT_GESAMTTYP
4        FINANZ_MINIMALIST
              ...         
80                PLZ8_HHZ
81                PLZ8_GBZ
82                  ARBEIT
83             ORTSGR_KLS9
84                RELAT_AB
Name: attribute, Length: 78, dtype: object 

Now I will print the missing values given in the Summary file for each column.

# give 'numeric' at the top of the print, don't sort the rest of the type values.
#custom_dict = {'numeric': 1}


with pd.option_context('display.max_rows', None,):
   #print(feat_info.sort_values(by=['type'], key=lambda x: x.map(custom_dict)))
   print(feat_info)

                attribute information_level         type missing_or_unknown
0                AGER_TYP            person  categorical             [-1,0]
1    ALTERSKATEGORIE_GROB            person      ordinal           [-1,0,9]
2               ANREDE_KZ            person  categorical             [-1,0]
3           CJT_GESAMTTYP            person  categorical                [0]
4       FINANZ_MINIMALIST            person      ordinal               [-1]
5           FINANZ_SPARER            person      ordinal               [-1]
6        FINANZ_VORSORGER            person      ordinal               [-1]
7          FINANZ_ANLEGER            person      ordinal               [-1]
8   FINANZ_UNAUFFAELLIGER            person      ordinal               [-1]
9        FINANZ_HAUSBAUER            person      ordinal               [-1]
10              FINANZTYP            person  categorical               [-1]
11            GEBURTSJAHR            person      numeric                [0]
12        GFK_URLAUBERTYP            person  categorical                 []
13       GREEN_AVANTGARDE            person  categorical                 []
14             HEALTH_TYP            person      ordinal             [-1,0]
15    LP_LEBENSPHASE_FEIN            person        mixed                [0]
16    LP_LEBENSPHASE_GROB            person        mixed                [0]
17        LP_FAMILIE_FEIN            person  categorical                [0]
18        LP_FAMILIE_GROB            person  categorical                [0]
19         LP_STATUS_FEIN            person  categorical                [0]
20         LP_STATUS_GROB            person  categorical                [0]
21       NATIONALITAET_KZ            person  categorical             [-1,0]
22  PRAEGENDE_JUGENDJAHRE            person        mixed             [-1,0]
23         RETOURTYP_BK_S            person      ordinal                [0]
24              SEMIO_SOZ            person      ordinal             [-1,9]
25              SEMIO_FAM            person      ordinal             [-1,9]
26              SEMIO_REL            person      ordinal             [-1,9]
27              SEMIO_MAT            person      ordinal             [-1,9]
28             SEMIO_VERT            person      ordinal             [-1,9]
29             SEMIO_LUST            person      ordinal             [-1,9]
30              SEMIO_ERL            person      ordinal             [-1,9]
31             SEMIO_KULT            person      ordinal             [-1,9]
32              SEMIO_RAT            person      ordinal             [-1,9]
33             SEMIO_KRIT            person      ordinal             [-1,9]
34              SEMIO_DOM            person      ordinal             [-1,9]
35             SEMIO_KAEM            person      ordinal             [-1,9]
36          SEMIO_PFLICHT            person      ordinal             [-1,9]
37            SEMIO_TRADV            person      ordinal             [-1,9]
38            SHOPPER_TYP            person  categorical               [-1]
39                SOHO_KZ            person  categorical               [-1]
40               TITEL_KZ            person  categorical             [-1,0]
41               VERS_TYP            person  categorical               [-1]
42               ZABEOTYP            person  categorical             [-1,9]
43               ALTER_HH         household     interval                [0]
44           ANZ_PERSONEN         household      numeric                 []
45              ANZ_TITEL         household      numeric                 []
46     HH_EINKOMMEN_SCORE         household      ordinal             [-1,0]
47           KK_KUNDENTYP         household  categorical               [-1]
48         W_KEIT_KIND_HH         household      ordinal             [-1,0]
49         WOHNDAUER_2008         household      ordinal             [-1,0]
50    ANZ_HAUSHALTE_AKTIV          building      numeric                [0]
51           ANZ_HH_TITEL          building      numeric                 []
52            GEBAEUDETYP          building  categorical             [-1,0]
53            KONSUMNAEHE          building      ordinal                 []
54       MIN_GEBAEUDEJAHR          building      numeric                [0]
55            OST_WEST_KZ          building  categorical               [-1]
56               WOHNLAGE          building        mixed               [-1]
57        CAMEO_DEUG_2015     microcell_rr4  categorical             [-1,X]
58         CAMEO_DEU_2015     microcell_rr4  categorical               [XX]
59        CAMEO_INTL_2015     microcell_rr4        mixed            [-1,XX]
60            KBA05_ANTG1     microcell_rr3      ordinal               [-1]
61            KBA05_ANTG2     microcell_rr3      ordinal               [-1]
62            KBA05_ANTG3     microcell_rr3      ordinal               [-1]
63            KBA05_ANTG4     microcell_rr3      ordinal               [-1]
64           KBA05_BAUMAX     microcell_rr3        mixed             [-1,0]
65              KBA05_GBZ     microcell_rr3      ordinal             [-1,0]
66               BALLRAUM          postcode      ordinal               [-1]
67               EWDICHTE          postcode      ordinal               [-1]
68             INNENSTADT          postcode      ordinal               [-1]
69     GEBAEUDETYP_RASTER        region_rr1      ordinal                 []
70                    KKK        region_rr1      ordinal             [-1,0]
71             MOBI_REGIO        region_rr1      ordinal                 []
72      ONLINE_AFFINITAET        region_rr1      ordinal                 []
73               REGIOTYP        region_rr1      ordinal             [-1,0]
74       KBA13_ANZAHL_PKW    macrocell_plz8      numeric                 []
75             PLZ8_ANTG1    macrocell_plz8      ordinal               [-1]
76             PLZ8_ANTG2    macrocell_plz8      ordinal               [-1]
77             PLZ8_ANTG3    macrocell_plz8      ordinal               [-1]
78             PLZ8_ANTG4    macrocell_plz8      ordinal               [-1]
79            PLZ8_BAUMAX    macrocell_plz8        mixed             [-1,0]
80               PLZ8_HHZ    macrocell_plz8      ordinal               [-1]
81               PLZ8_GBZ    macrocell_plz8      ordinal               [-1]
82                 ARBEIT         community      ordinal             [-1,9]
83            ORTSGR_KLS9         community      ordinal             [-1,0]
84               RELAT_AB         community      ordinal             [-1,9]

Looking at the pie charts, I HAVE NOTICED other missing values or incorrect values than those marked in the Summary file! For example: for columns like KBA05_ANTG1 to KBA05_ANTG4 there should be "-1" value for missing_unknown, as it states in summary above. Yet, looking at the pie chart for those columns, we can't see such a value, but instead "null" value exists. In this particular case it's not a problem, because I'll overwrite every missing_unknown value to NaN anyway. However, as they say, caution is never enough and I feel more confident after doing my own analysis.

# For columns "b)" (numeric) I'll show histograms.

# making list of names of "b)" group columns
list_of_numeric_cols = feat_info[feat_info['type']=='numeric']['attribute']
print(list_of_numeric_cols, '\n')

for col_name in list_of_numeric_cols: 

    fig = px.histogram(azdias, x=col_name)
    fig.show()
    

11            GEBURTSJAHR
44           ANZ_PERSONEN
45              ANZ_TITEL
50    ANZ_HAUSHALTE_AKTIV
51           ANZ_HH_TITEL
54       MIN_GEBAEUDEJAHR
74       KBA13_ANZAHL_PKW
Name: attribute, dtype: object 

'''
# Identify missing or unknown data values and convert them to NaNs.

def parser(string):
    list_of_values = []
    
    if string == "[]":
        return list_of_values

    string = string.strip(' ').strip('')
    string_without_brackets = string.lstrip("[").rstrip("]")
    list_of_string_values = string_without_brackets.split(",")
    
    print("List: {}; Len {}".format(list_of_string_values, len(list_of_string_values)))

    for i in list_of_string_values:
        if i == 'X' or i == 'XX':
            list_of_values.append(str(i))
        else:
            list_of_values.append(int(i))
        

    return list_of_values
'''

# Identify missing or unknown data values and convert them to NaNs.

def parser(string):
    list_of_values = []
    
    if string == "[]":
        return list_of_values

    string = string.strip(' ').strip('')
    string_without_brackets = string.lstrip("[").rstrip("]")
    list_of_string_values = string_without_brackets.split(",")
    
    print("List: {}; Len {}".format(list_of_string_values, len(list_of_string_values)))

    for i in list_of_string_values:
        if i == 'X' or i == 'XX':
            list_of_values.append(str(i))
        else:
            list_of_values.append(int(i))
        
    return list_of_values

for column in azdias:
    str_of_unknown = feat_info[feat_info['attribute']==column]['missing_or_unknown'].values[0]
    list_of_unknown = parser(str_of_unknown)
    print(column)
    
    for val in list_of_unknown:
        print(val)
        azdias[column] = azdias[column].replace(val, np.nan)
        print(azdias[column].unique())




        

List: ['-1', '0']; Len 2
AGER_TYP
-1
[nan  2.  3.  0.  1.]
0
[nan  2.  3.  1.]
List: ['-1', '0', '9']; Len 3
ALTERSKATEGORIE_GROB
-1
[2 1 3 4 9]
0
[2 1 3 4 9]
9
[ 2.  1.  3.  4. nan]
List: ['-1', '0']; Len 2
ANREDE_KZ
-1
[1 2]
0
[1 2]
List: ['0']; Len 1
CJT_GESAMTTYP
0
[ 2.  5.  3.  4.  1.  6. nan]
List: ['-1']; Len 1
FINANZ_MINIMALIST
-1
[3 1 4 2 5]
List: ['-1']; Len 1
FINANZ_SPARER
-1
[4 5 2 3 1]
List: ['-1']; Len 1
FINANZ_VORSORGER
-1
[3 2 1 5 4]
List: ['-1']; Len 1
FINANZ_ANLEGER
-1
[5 2 1 4 3]
List: ['-1']; Len 1
FINANZ_UNAUFFAELLIGER
-1
[5 4 3 1 2]
List: ['-1']; Len 1
FINANZ_HAUSBAUER
-1
[3 5 2 4 1]
List: ['-1']; Len 1
FINANZTYP
-1
[4 1 6 5 2 3]
List: ['0']; Len 1
GEBURTSJAHR
0
[  nan 1996. 1979. 1957. 1963. 1943. 1964. 1974. 1975. 1941. 1993. 1946.
 1988. 1991. 1966. 1949. 1982. 1972. 1990. 1985. 1989. 1948. 1935. 1997.
 1994. 1987. 2002. 1950. 1965. 1969. 1978. 1958. 1944. 1967. 1977. 1960.
 1970. 1973. 1971. 1937. 1947. 1926. 1954. 1981. 1962. 1992. 2008. 1976.
 1928. 1927. 1968. 1940. 1952. 1961. 2006. 1938. 1945. 1942. 1984. 1930.
 1980. 1959. 1986. 1931. 1955. 1983. 2010. 1951. 1921. 1929. 1939. 2001.
 1995. 1956. 1953. 2012. 1934. 1913. 1914. 1922. 2004. 1936. 2009. 1915.
 2003. 2013. 1998. 2007. 1932. 2005. 1923. 1925. 2015. 1933. 2017. 2000.
 1999. 2014. 1920. 1918. 1924. 2016. 2011. 1908. 1919. 1916. 1917. 1912.
 1911. 1909. 1910. 1904. 1906. 1905. 1900. 1907. 1902.]
GFK_URLAUBERTYP
GREEN_AVANTGARDE
List: ['-1', '0']; Len 2
HEALTH_TYP
-1
[nan  3.  2.  1.]
0
[nan  3.  2.  1.]
List: ['0']; Len 1
LP_LEBENSPHASE_FEIN
0
[15. 21.  3. nan 32.  8.  2.  5. 10.  4.  6. 23. 12. 20.  1. 11. 25. 13.
  7. 18. 31. 19. 38. 35. 30. 22. 14. 33. 29. 24. 28. 37. 26. 39. 27. 36.
  9. 34. 40. 16. 17.]
List: ['0']; Len 1
LP_LEBENSPHASE_GROB
0
[ 4.  6.  1. nan 10.  2.  3.  5.  7. 12. 11.  9.  8.]
List: ['0']; Len 1
LP_FAMILIE_FEIN
0
[ 2.  5.  1. nan 10.  7. 11.  3.  8.  4.  6.  9.]
List: ['0']; Len 1
LP_FAMILIE_GROB
0
[ 2.  3.  1. nan  5.  4.]
List: ['0']; Len 1
LP_STATUS_FEIN
0
[ 1.  2.  3.  9.  4. 10.  5.  8.  6.  7. nan]
List: ['0']; Len 1
LP_STATUS_GROB
0
[ 1.  2.  4.  5.  3. nan]
List: ['-1', '0']; Len 2
NATIONALITAET_KZ
-1
[0 1 3 2]
0
[nan  1.  3.  2.]
List: ['-1', '0']; Len 2
PRAEGENDE_JUGENDJAHRE
-1
[ 0 14 15  8  3 10 11  5  9  6  4  2  1 12 13  7]
0
[nan 14. 15.  8.  3. 10. 11.  5.  9.  6.  4.  2.  1. 12. 13.  7.]
List: ['0']; Len 1
RETOURTYP_BK_S
0
[ 5.  1.  3.  2.  4. nan]
List: ['-1', '9']; Len 2
SEMIO_SOZ
-1
[2 5 4 6 7 3 1]
9
[2 5 4 6 7 3 1]
List: ['-1', '9']; Len 2
SEMIO_FAM
-1
[6 4 1 5 7 2 3]
9
[6 4 1 5 7 2 3]
List: ['-1', '9']; Len 2
SEMIO_REL
-1
[7 4 3 2 5 1 6]
9
[7 4 3 2 5 1 6]
List: ['-1', '9']; Len 2
SEMIO_MAT
-1
[5 3 1 2 4 7 6]
9
[5 3 1 2 4 7 6]
List: ['-1', '9']; Len 2
SEMIO_VERT
-1
[1 4 7 2 6 5 3]
9
[1 4 7 2 6 5 3]
List: ['-1', '9']; Len 2
SEMIO_LUST
-1
[5 2 4 6 7 3 1]
9
[5 2 4 6 7 3 1]
List: ['-1', '9']; Len 2
SEMIO_ERL
-1
[3 2 6 7 4 5 1]
9
[3 2 6 7 4 5 1]
List: ['-1', '9']; Len 2
SEMIO_KULT
-1
[3 4 6 5 7 1 2]
9
[3 4 6 5 7 1 2]
List: ['-1', '9']; Len 2
SEMIO_RAT
-1
[4 6 3 2 7 5 1]
9
[4 6 3 2 7 5 1]
List: ['-1', '9']; Len 2
SEMIO_KRIT
-1
[7 4 3 1 5 6 2]
9
[7 4 3 1 5 6 2]
List: ['-1', '9']; Len 2
SEMIO_DOM
-1
[6 7 4 2 1 5 3]
9
[6 7 4 2 1 5 3]
List: ['-1', '9']; Len 2
SEMIO_KAEM
-1
[6 4 7 5 2 3 1]
9
[6 4 7 5 2 3 1]
List: ['-1', '9']; Len 2
SEMIO_PFLICHT
-1
[5 7 3 4 1 6 2]
9
[5 7 3 4 1 6 2]
List: ['-1', '9']; Len 2
SEMIO_TRADV
-1
[3 6 4 2 7 5 1]
9
[3 6 4 2 7 5 1]
List: ['-1']; Len 1
SHOPPER_TYP
-1
[nan  3.  2.  1.  0.]
List: ['-1']; Len 1
SOHO_KZ
-1
[nan  1.  0.]
List: ['-1', '0']; Len 2
TITEL_KZ
-1
[nan  0.  4.  1.  3.  5.  2.]
0
[nan  4.  1.  3.  5.  2.]
List: ['-1']; Len 1
VERS_TYP
-1
[nan  2.  1.]
List: ['-1', '9']; Len 2
ZABEOTYP
-1
[3 5 4 1 6 2]
9
[3 5 4 1 6 2]
List: ['0']; Len 1
ALTER_HH
0
[nan 17. 13. 20. 10. 14. 16. 21. 11. 19. 15.  9. 18.  8.  7. 12.  4.  3.
  6.  5.  2.  1.]
ANZ_PERSONEN
ANZ_TITEL
List: ['-1', '0']; Len 2
HH_EINKOMMEN_SCORE
-1
[ 2.  6.  4.  1.  5.  3. nan]
0
[ 2.  6.  4.  1.  5.  3. nan]
List: ['-1']; Len 1
KK_KUNDENTYP
-1
[nan  1.  3.  6.  4.  2.  5.]
List: ['-1', '0']; Len 2
W_KEIT_KIND_HH
-1
[nan  3.  2.  6.  5.  4.  0.  1.]
0
[nan  3.  2.  6.  5.  4.  1.]
List: ['-1', '0']; Len 2
WOHNDAUER_2008
-1
[nan  9.  8.  3.  4.  5.  6.  2.  7.  1.]
0
[nan  9.  8.  3.  4.  5.  6.  2.  7.  1.]
List: ['0']; Len 1
ANZ_HAUSHALTE_AKTIV
0
[ nan  11.  10.   1.   3.   5.   4.   6.   2.   9.  14.  13.   8.   7.
  44.  25.  36.  16. 272.  33.  15.  12.  22.  21.  20.  67.  23.  18.
  27.  24.  31.  39.  19.  30.  17.  29.  61.  54.  88.  49.  60.  38.
  34.  48.  35.  62.  26.  45.  42.  28.  77.  47.  50. 236. 137.  43.
  52.  41.  76.  63.  32.  40.  37.  59.  53.  57. 118.  66.  95.  78.
  70.  46. 185.  73. 154.  80. 102.  92. 126.  94.  71.  68.  81.  69.
 121. 258.  58.  86. 124.  85.  90. 128.  51.  72.  64.  56.  83. 150.
 259.  91. 160. 134. 169. 246.  82. 114. 183. 116.  74. 143.  99. 216.
  98. 125. 112. 123.  55. 120.  75. 141. 103.  87.  89. 223.  93. 179.
 135. 149.  97. 101. 110.  79. 111. 109.  65.  84. 367. 162. 195. 107.
 119. 331. 122. 161. 157. 175. 177. 136. 104. 316. 127. 100. 180. 155.
 133. 178. 176. 108. 353. 131. 130.  96. 256. 117. 146. 211. 199. 140.
 200. 106. 285. 263. 314. 253. 163. 377. 171. 148. 191. 243. 153. 145.
 174. 172. 190. 142. 317. 244. 189. 286. 105. 170. 215. 321. 113. 115.
 156. 129. 242. 132. 280. 159. 138. 158. 267. 151. 164. 187. 206. 266.
 265. 209. 196. 139. 193. 184. 210. 260. 366. 194. 318. 344. 277. 595.
 197. 247. 165. 283. 326. 181. 186. 438. 222. 144. 378. 237. 152. 208.
 274. 430. 202. 214. 231. 219. 225. 230. 198. 224. 333. 168. 221. 523.
 254. 305. 166. 252. 445. 147. 301. 213. 536. 220. 249. 290. 515. 276.
 241. 204. 250. 218. 226. 201. 240. 404. 348. 328. 228. 304. 167. 346.
 205. 207. 293. 255. 307. 414. 311. 347. 379. 395. 229. 232.]
ANZ_HH_TITEL
List: ['-1', '0']; Len 2
GEBAEUDETYP
-1
[nan  8.  1.  3.  2.  6.  4.  5.]
0
[nan  8.  1.  3.  2.  6.  4.  5.]
KONSUMNAEHE
List: ['0']; Len 1
MIN_GEBAEUDEJAHR
0
[  nan 1992. 1997. 2005. 2009. 1994. 1996. 2002. 2015. 1991. 1993. 1995.
 2003. 2014. 2008. 2006. 2000. 1990. 2004. 1999. 1998. 2001. 2007. 2013.
 1989. 2011. 2012. 2010. 1987. 1988. 1985. 2016. 1986.]
List: ['-1']; Len 1
OST_WEST_KZ
-1
[nan 'W' 'O']
List: ['-1']; Len 1
WOHNLAGE
-1
[nan  4.  2.  7.  3.  5.  1.  8.  0.]
List: ['-1', 'X']; Len 2
CAMEO_DEUG_2015
-1
[nan '8' '4' '2' '6' '1' '9' '5' '7' '3' 'X']
X
[nan '8' '4' '2' '6' '1' '9' '5' '7' '3']
List: ['XX']; Len 1
CAMEO_DEU_2015
XX
[nan '8A' '4C' '2A' '6B' '8C' '4A' '2D' '1A' '1E' '9D' '5C' '8B' '7A' '5D'
 '9E' '9B' '1B' '3D' '4E' '4B' '3C' '5A' '7B' '9A' '6D' '6E' '2C' '7C'
 '9C' '7D' '5E' '1D' '8D' '6C' '6A' '5B' '4D' '3A' '2B' '7E' '3B' '6F'
 '5F' '1C']
List: ['-1', 'XX']; Len 2
CAMEO_INTL_2015
-1
[nan '51' '24' '12' '43' '54' '22' '14' '13' '15' '33' '41' '34' '55' '25'
 '23' '31' '52' '35' '45' '44' '32' 'XX']
XX
[nan '51' '24' '12' '43' '54' '22' '14' '13' '15' '33' '41' '34' '55' '25'
 '23' '31' '52' '35' '45' '44' '32']
List: ['-1']; Len 1
KBA05_ANTG1
-1
[nan  0.  1.  4.  2.  3.]
List: ['-1']; Len 1
KBA05_ANTG2
-1
[nan  0.  3.  1.  4.  2.]
List: ['-1']; Len 1
KBA05_ANTG3
-1
[nan  0.  1.  2.  3.]
List: ['-1']; Len 1
KBA05_ANTG4
-1
[nan  2.  0.  1.]
List: ['-1', '0']; Len 2
KBA05_BAUMAX
-1
[nan  5.  0.  1.  2.  3.  4.]
0
[nan  5.  1.  2.  3.  4.]
List: ['-1', '0']; Len 2
KBA05_GBZ
-1
[nan  1.  3.  4.  5.  2.]
0
[nan  1.  3.  4.  5.  2.]
List: ['-1']; Len 1
BALLRAUM
-1
[nan  6.  2.  4.  3.  7.  1.  5.]
List: ['-1']; Len 1
EWDICHTE
-1
[nan  3.  4.  2.  5.  6.  1.]
List: ['-1']; Len 1
INNENSTADT
-1
[nan  8.  4.  6.  1.  7.  3.  2.  5.]
GEBAEUDETYP_RASTER
List: ['-1', '0']; Len 2
KKK
-1
[nan  2.  0.  3.  4.  1.]
0
[nan  2.  3.  4.  1.]
MOBI_REGIO
ONLINE_AFFINITAET
List: ['-1', '0']; Len 2
REGIOTYP
-1
[nan  3.  2.  0.  5.  1.  7.  6.  4.]
0
[nan  3.  2.  5.  1.  7.  6.  4.]
KBA13_ANZAHL_PKW
List: ['-1']; Len 1
PLZ8_ANTG1
-1
[nan  2.  3.  1.  4.  0.]
List: ['-1']; Len 1
PLZ8_ANTG2
-1
[nan  3.  2.  4.  1.  0.]
List: ['-1']; Len 1
PLZ8_ANTG3
-1
[nan  2.  1.  3.  0.]
List: ['-1']; Len 1
PLZ8_ANTG4
-1
[nan  1.  0.  2.]
List: ['-1', '0']; Len 2
PLZ8_BAUMAX
-1
[nan  1.  2.  4.  5.  3.]
0
[nan  1.  2.  4.  5.  3.]
List: ['-1']; Len 1
PLZ8_HHZ
-1
[nan  5.  4.  3.  2.  1.]
List: ['-1']; Len 1
PLZ8_GBZ
-1
[nan  4.  3.  5.  2.  1.]
List: ['-1', '9']; Len 2
ARBEIT
-1
[nan  3.  2.  4.  1.  5.  9.]
9
[nan  3.  2.  4.  1.  5.]
List: ['-1', '0']; Len 2
ORTSGR_KLS9
-1
[nan  5.  3.  6.  4.  8.  2.  7.  9.  1.  0.]
0
[nan  5.  3.  6.  4.  8.  2.  7.  9.  1.]
List: ['-1', '9']; Len 2
RELAT_AB
-1
[nan  4.  2.  3.  5.  1.  9.]
9
[nan  4.  2.  3.  5.  1.]

Now I'll show the diagrams after missing/unknown data parsing to NaNs.

'''
for i, col_name in zip(range(4), list_of_categorical_cols): 
    fig = px.pie(azdias[col_name].value_counts(dropna=False), values=col_name, 
             names = azdias[col_name].value_counts(dropna=False).index, title = col_name, template='ggplot2') 
    
    fig.show()
'''

for col_name in list_of_categorical_cols: 
    fig = px.pie(azdias[col_name].value_counts(dropna=False), values=col_name, 
             names = azdias[col_name].value_counts(dropna=False).index, title = col_name, template='ggplot2') 
    
    fig.show()

Step 1.1.2: Assess Missing Data in Each Column

How much missing data is present in each column? There are a few columns that are outliers in terms of the proportion of values that are missing. You will want to use matplotlib's hist() function to visualize the distribution of missing value counts to find these columns. Identify and document these columns. While some of these columns might have justifications for keeping or re-encoding the data, for this project you should just remove them from the dataframe. (Feel free to make remarks about these outlier columns in the discussion, however!)

For the remaining features, are there any patterns in which columns have, or share, missing data?

# Perform an assessment of how much missing data there is in each column of the
# dataset.

percent_of_missing_in_columns = []
names_of_columns = []

for c in azdias.columns:
    names_of_columns.append(c)
    percent_of_missing_in_columns.append((azdias[c].isna().sum() / azdias.shape[0]) * 100)
        

# Investigate patterns in the amount of missing data in each column.

fig = plt.figure(figsize=(20,8))
plt.bar(names_of_columns, percent_of_missing_in_columns)
plt.ylabel("Percent of NaN values per column")
plt.xticks(rotation='vertical')
plt.axhline(y=20, color='r', linestyle='-')
plt.axhline(y=8, color='g', linestyle='-')
plt.grid(axis="y")
plt.show()

# Remove the outlier columns from the dataset. (You'll perform other data
# engineering tasks such as re-encoding and imputation later.)

for col, percent in zip(names_of_columns, percent_of_missing_in_columns):
    if percent >= 20 :
        azdias.drop(col, axis=1, inplace=True)
        print(col)
        
        
print('\n', azdias.shape)

AGER_TYP
GEBURTSJAHR
TITEL_KZ
ALTER_HH
KK_KUNDENTYP
KBA05_BAUMAX

 (891221, 79)

Discussion 1.1.2: Assess Missing Data in Each Column

After drawing the diagram above, at first glance, I see 3 ranges of the number of missing values:

• outliers, i.e. columns where the missing values are above 20% (6 columns),
• the range of 8-17% of which is the majority,
• and the range below 4% of which is about 1 / 3.

Due to the significant lack of data, I removed the following columns:

• 'AGER_TYP',
• 'GEBURTSJAHR',
• 'TITEL_KZ',
• 'ALTER_HH',
• 'KK_KUNDENTYP',
• 'KBA05_BAUMAX'

Step 1.1.3: Assess Missing Data in Each Row

Now, you'll perform a similar assessment for the rows of the dataset. How much data is missing in each row? As with the columns, you should see some groups of points that have a very different numbers of missing values. Divide the data into two subsets: one for data points that are above some threshold for missing values, and a second subset for points below that threshold.

In order to know what to do with the outlier rows, we should see if the distribution of data values on columns that are not missing data (or are missing very little data) are similar or different between the two groups. Select at least five of these columns and compare the distribution of values.

• You can use seaborn's countplot() function to create a bar chart of code frequencies and matplotlib's subplot() function to put bar charts for the two subplots side by side.
• To reduce repeated code, you might want to write a function that can perform this comparison, taking as one of its arguments a column to be compared.

Depending on what you observe in your comparison, this will have implications on how you approach your conclusions later in the analysis. If the distributions of non-missing features look similar between the data with many missing values and the data with few or no missing values, then we could argue that simply dropping those points from the analysis won't present a major issue. On the other hand, if the data with many missing values looks very different from the data with few or no missing values, then we should make a note on those data as special. We'll revisit these data later on. Either way, you should continue your analysis for now using just the subset of the data with few or no missing values.

# How much data is missing in each row of the dataset?
nan_in_rows = azdias.isnull().sum(axis=1)
plt.hist(nan_in_rows, bins=50)
plt.xlabel('Number of NaN per row')
plt.ylabel('Number of rows')
plt.show()

print("All rows number: ", azdias.shape[0])
print("Rows having at least one NaN value: ", azdias.shape[0]- azdias.dropna().shape[0])
print("Rows having all values (no NaN): {}, {:.2f}%".format(azdias.dropna().shape[0], azdias.dropna().shape[0] * 100 / azdias.shape[0] ))
print("Max number of missing values in a row: ", azdias.isna().sum(axis=1).max())

All rows number:  891221
Rows having at least one NaN value:  268012
Rows having all values (no NaN): 623209, 69.93%
Max number of missing values in a row:  49

This is an optimistic info to see, that almost 70% of rows are full (no missing values).

# Write code to divide the data into two subsets based on the number of missing
# values in each row.

treshold = 9

rows_below_treshold = azdias[nan_in_rows <= treshold]
rows_above_treshold = azdias[nan_in_rows > treshold]

print(rows_above_treshold.shape)
print(rows_below_treshold.shape)

(116478, 79)
(774743, 79)

# Compare the distribution of values for at least five columns where there are
# no or few missing values, between the two subsets.

fig, axs = plt.subplots(nrows=10, ncols=2, figsize=(30, 80))

n = 0
for column in azdias.columns[10:20]:
    sns.countplot(x=rows_below_treshold.loc[:, column], ax=axs[n,0])
    axs[n,0].set_title('rows_below_treshold')
    sns.countplot(x=rows_above_treshold.loc[:, column], ax=axs[n,1])
    axs[n,1].set_title('rows_above_treshold')
    n += 1

Discussion 1.1.3: Assess Missing Data in Each Row

For the following pairs, I find these distributions rather similar to each other. Moreover, the part of the data with rows below the threshold point contains more data than the part with rows above the threshold point. The data with lots of missing values are NOT qualitatively different (in most cases) from data with few or no missing values. In my opinion, rows with more than 9 missing data can be rejected before further analysis.

print("I'm dropping {:.2f} % of low quality data.".format(rows_above_treshold.shape[0] * 100.0/ azdias.shape[0]))

I'm dropping 13.07 % of low quality data.

# working on that df from now on
df_few_missing = rows_below_treshold.copy(deep=True)

Step 1.2: Select and Re-Encode Features

Checking for missing data isn't the only way in which you can prepare a dataset for analysis. Since the unsupervised learning techniques to be used will only work on data that is encoded numerically, you need to make a few encoding changes or additional assumptions to be able to make progress. In addition, while almost all of the values in the dataset are encoded using numbers, not all of them represent numeric values. Check the third column of the feature summary (feat_info) for a summary of types of measurement.

• For numeric and interval data, these features can be kept without changes.
• Most of the variables in the dataset are ordinal in nature. While ordinal values may technically be non-linear in spacing, make the simplifying assumption that the ordinal variables can be treated as being interval in nature (that is, kept without any changes).
• Special handling may be necessary for the remaining two variable types: categorical, and 'mixed'.

In the first two parts of this sub-step, you will perform an investigation of the categorical and mixed-type features and make a decision on each of them, whether you will keep, drop, or re-encode each. Then, in the last part, you will create a new data frame with only the selected and engineered columns.

Data wrangling is often the trickiest part of the data analysis process, and there's a lot of it to be done here. But stick with it: once you're done with this step, you'll be ready to get to the machine learning parts of the project!

# How many features are there of each data type?

feat_info.groupby(['type'])['attribute'].count()

type
categorical    21
interval        1
mixed           7
numeric         7
ordinal        49
Name: attribute, dtype: int64

Step 1.2.1: Re-Encode Categorical Features

For categorical data, you would ordinarily need to encode the levels as dummy variables. Depending on the number of categories, perform one of the following:

• For binary (two-level) categoricals that take numeric values, you can keep them without needing to do anything.
• There is one binary variable that takes on non-numeric values. For this one, you need to re-encode the values as numbers or create a dummy variable.
• For multi-level categoricals (three or more values), you can choose to encode the values using multiple dummy variables (e.g. via OneHotEncoder), or (to keep things straightforward) just drop them from the analysis. As always, document your choices in the Discussion section.

# Assess categorical variables: which are binary, which are multi-level, and
# which one needs to be re-encoded?

categorical_cols = feat_info[feat_info['type'] == 'categorical']['attribute'].values
#print(categorical_cols)

cat_cols_in_new_df = [x for x in categorical_cols if x in df_few_missing.columns]
print("All categorical in new dataframe: \n", cat_cols_in_new_df, "\n\n")


binary = [x for x in cat_cols_in_new_df if df_few_missing[x].nunique()==2]
multilevel_categorical = [x for x in cat_cols_in_new_df if df_few_missing[x].nunique()>2]
#print(df_few_missing[cat_cols_in_new_df].nunique())

print("Binary: \n", binary, "\n")

print("Multilevel: \n", multilevel_categorical)

All categorical in new dataframe: 
 ['ANREDE_KZ', 'CJT_GESAMTTYP', 'FINANZTYP', 'GFK_URLAUBERTYP', 'GREEN_AVANTGARDE', 'LP_FAMILIE_FEIN', 'LP_FAMILIE_GROB', 'LP_STATUS_FEIN', 'LP_STATUS_GROB', 'NATIONALITAET_KZ', 'SHOPPER_TYP', 'SOHO_KZ', 'VERS_TYP', 'ZABEOTYP', 'GEBAEUDETYP', 'OST_WEST_KZ', 'CAMEO_DEUG_2015', 'CAMEO_DEU_2015'] 


Binary: 
 ['ANREDE_KZ', 'GREEN_AVANTGARDE', 'SOHO_KZ', 'VERS_TYP', 'OST_WEST_KZ'] 

Multilevel: 
 ['CJT_GESAMTTYP', 'FINANZTYP', 'GFK_URLAUBERTYP', 'LP_FAMILIE_FEIN', 'LP_FAMILIE_GROB', 'LP_STATUS_FEIN', 'LP_STATUS_GROB', 'NATIONALITAET_KZ', 'SHOPPER_TYP', 'ZABEOTYP', 'GEBAEUDETYP', 'CAMEO_DEUG_2015', 'CAMEO_DEU_2015']

# Converting non-numeric values with binary values

for b in binary:
    print(b, df_few_missing[b].unique())
    

ANREDE_KZ [2 1]
GREEN_AVANTGARDE [0 1]
SOHO_KZ [1. 0.]
VERS_TYP [ 2.  1. nan]
OST_WEST_KZ ['W' 'O']

#df_few_missing.loc[:, 'ANREDE_KZ'].replace({'1':0, '2':1}, inplace=True)
#df_few_missing.loc[:, 'GREEN_AVANTGARDE'].replace({'0':0, '1':1}, inplace=True)
#df_few_missing.loc[:, 'SOHO_KZ'].replace({0.:0, 1.:1}, inplace=True)
#df_few_missing.loc[:, 'VERS_TYP'].replace({1.:0, 2.:1}, inplace=True)
df_few_missing.loc[:, 'OST_WEST_KZ'].replace({'W':0, 'O':1}, inplace=True)

for b in binary:
    print(b, df_few_missing[b].unique())

ANREDE_KZ [2 1]
GREEN_AVANTGARDE [0 1]
SOHO_KZ [1. 0.]
VERS_TYP [ 2.  1. nan]
OST_WEST_KZ [0 1]

# Before doing dummy variables I need to clean the data from NaNs.
# For now I'm replacing null with mode (common value), maybe I'll change that to KNNImputer later


num_of_categorical_cols_with_nan = 0
for col in cat_cols_in_new_df:
    if df_few_missing[col].isnull().values.any() == True: 
        num_of_categorical_cols_with_nan += 1
        print(col, df_few_missing[col].unique())
        df_few_missing[col] = df_few_missing[col].fillna(df_few_missing[col].mode()[0])
        
        
print(num_of_categorical_cols_with_nan)

LP_FAMILIE_FEIN [ 5.  1. nan 10.  2.  7. 11.  8.  4.  6.  9.  3.]
LP_FAMILIE_GROB [ 3.  1. nan  5.  2.  4.]
NATIONALITAET_KZ [ 1.  3.  2. nan]
SHOPPER_TYP [ 3.  2.  1.  0. nan]
VERS_TYP [ 2.  1. nan]
CAMEO_DEUG_2015 ['8' '4' '2' '6' '1' '9' '5' '7' '3' nan]
CAMEO_DEU_2015 ['8A' '4C' '2A' '6B' '8C' '4A' '2D' '1A' '1E' '9D' '5C' '8B' '7A' '5D'
 '9E' '9B' '1B' '3D' nan '4E' '4B' '3C' '5A' '7B' '9A' '6D' '6E' '2C' '7C'
 '9C' '7D' '5E' '1D' '8D' '6C' '6A' '5B' '4D' '3A' '2B' '7E' '3B' '6F'
 '5F' '1C']
7

# Re-encode categorical variable(s) to be kept in the analysis.


df_few_missing = pd.get_dummies(df_few_missing, columns=multilevel_categorical)

df_few_missing.head()

	ALTERSKATEGORIE_GROB	ANREDE_KZ	FINANZ_MINIMALIST	FINANZ_SPARER	FINANZ_VORSORGER	FINANZ_ANLEGER	FINANZ_UNAUFFAELLIGER	FINANZ_HAUSBAUER	GREEN_AVANTGARDE	HEALTH_TYP	...	CAMEO_DEU_2015_7E	CAMEO_DEU_2015_8A	CAMEO_DEU_2015_8B	CAMEO_DEU_2015_8C	CAMEO_DEU_2015_8D	CAMEO_DEU_2015_9A	CAMEO_DEU_2015_9B	CAMEO_DEU_2015_9C	CAMEO_DEU_2015_9D	CAMEO_DEU_2015_9E
1	1.0	2	1	5	2	5	4	5	0	3.0	...	0	1	0	0	0	0	0	0	0	0
2	3.0	2	1	4	1	2	3	5	1	3.0	...	0	0	0	0	0	0	0	0	0	0
3	4.0	2	4	2	5	2	1	2	0	2.0	...	0	0	0	0	0	0	0	0	0	0
4	3.0	1	4	3	4	1	3	2	0	3.0	...	0	0	0	0	0	0	0	0	0	0
5	1.0	2	3	1	5	2	2	5	0	3.0	...	0	0	0	1	0	0	0	0	0	0

5 rows × 194 columns

Discussion 1.2.1: Re-Encode Categorical Features

• I converted non-numeric values with binary values for 'OST_WEST_KZ' column
• I leaved as they are categorical columns with numeric binary values
• To do the below point, I got rid of missing values (in categorical columns only, for now)
• I converted the rest of multilevel categorical columns to one hot encoded columns.

Step 1.2.2: Engineer Mixed-Type Features

There are a handful of features that are marked as "mixed" in the feature summary that require special treatment in order to be included in the analysis. There are two in particular that deserve attention; the handling of the rest are up to your own choices:

• "PRAEGENDE_JUGENDJAHRE" combines information on three dimensions: generation by decade, movement (mainstream vs. avantgarde), and nation (east vs. west). While there aren't enough levels to disentangle east from west, you should create two new variables to capture the other two dimensions: an interval-type variable for decade, and a binary variable for movement.
• "CAMEO_INTL_2015" combines information on two axes: wealth and life stage. Break up the two-digit codes by their 'tens'-place and 'ones'-place digits into two new ordinal variables (which, for the purposes of this project, is equivalent to just treating them as their raw numeric values).
• If you decide to keep or engineer new features around the other mixed-type features, make sure you note your steps in the Discussion section.

Be sure to check Data_Dictionary.md for the details needed to finish these tasks.

# Investigate "PRAEGENDE_JUGENDJAHRE" and engineer two new variables.

'''
decade mapping:
 1: 40s
 2: 50s
 3: 60s
 4: 70s
 5: 80s
 6: 90s
 
 movement mapping:
 0: "Mainstream"
 1: "Avantgarde"
'''

PRAEGENDE_JUGENDJAHRE_decade = {1: 1 , 2: 1 , 3: 2 ,  4: 2 ,  5: 3 ,  6: 3 ,  7: 3 ,  8: 4 ,  9: 4 ,   10: 5 ,  11: 5 ,  12: 5 , 13: 5 ,  14: 6 , 15: 6 }                            
PRAEGENDE_JUGENDJAHRE_movement = {1: 0 , 2: 1 ,  3: 0 , 4: 1 , 5: 0 , 6: 1 ,  7: 1 ,  8: 0 ,  9: 1 ,  10: 0 , 11: 1 , 12: 0 , 13: 1 , 14: 0 , 15: 1 }

# Investigate "CAMEO_INTL_2015" and engineer two new variables.

CAMEO_INTL_2015_wealth = dict.fromkeys(["11", "12", "13", "14", "15"], 1)
CAMEO_INTL_2015_wealth.update(dict.fromkeys(["21", "22", "23", "24", "25"], 2))
CAMEO_INTL_2015_wealth.update(dict.fromkeys(["31", "32", "33", "34", "35"], 3))
CAMEO_INTL_2015_wealth.update(dict.fromkeys(["41", "42", "43", "44", "45"], 4))
CAMEO_INTL_2015_wealth.update(dict.fromkeys(["51", "52", "53", "54", "55"], 5))

CAMEO_INTL_2015_lifeStage = dict.fromkeys(["11", "21", "31", "41", "51"], 1)
CAMEO_INTL_2015_lifeStage.update(dict.fromkeys(["12", "22", "32", "42", "52"], 2))
CAMEO_INTL_2015_lifeStage.update(dict.fromkeys(["13", "23", "33", "43", "53"], 3))
CAMEO_INTL_2015_lifeStage.update(dict.fromkeys(["14", "24", "34", "44", "54"], 4))
CAMEO_INTL_2015_lifeStage.update(dict.fromkeys(["15", "25", "35", "45", "55"], 5))

# Putting new columns to the dataframe (new values are replaced on basis of old values from columns PRAEGENDE_JUGENDJAHRE and CAMEO_INTL_2015)

df_few_missing['PJ_DECADE'] = df_few_missing['PRAEGENDE_JUGENDJAHRE']
df_few_missing['PJ_DECADE'].replace(PRAEGENDE_JUGENDJAHRE_decade, inplace=True)
df_few_missing['PJ_MOVEMENT'] = df_few_missing['PRAEGENDE_JUGENDJAHRE']
df_few_missing['PJ_MOVEMENT'].replace(PRAEGENDE_JUGENDJAHRE_movement, inplace=True)

df_few_missing['CI_WEALTH'] = df_few_missing['CAMEO_INTL_2015']
df_few_missing['CI_WEALTH'].replace(CAMEO_INTL_2015_wealth, inplace=True)
df_few_missing['CI_LIFESTAGE'] = df_few_missing['CAMEO_INTL_2015']
df_few_missing['CI_LIFESTAGE'].replace(CAMEO_INTL_2015_lifeStage, inplace=True)

Discussion 1.2.2: Engineer Mixed-Type Features

I replaced old mixed columns: PRAEGENDE_JUGENDJAHRE and CAMEO_INTL_2015 with new ones, splitting data they described into 4 new columns: PJ_DECADE, PJ_MOVEMENT, CI_WEALTH, CI_LIFESTAGE.

I dropped LP_LEBENSPHASE_FEIN and LP_LEBENSPHASE_GROB, because it's hard to split the features they describe.

The rest of mixed columns (WOHNLAGE, PLZ8_BAUMAX) I treat as an ordinal columns, because they describes single features, they are not really 'mixed' features in my opinion.

Step 1.2.3: Complete Feature Selection

In order to finish this step up, you need to make sure that your data frame now only has the columns that you want to keep. To summarize, the dataframe should consist of the following:

• All numeric, interval, and ordinal type columns from the original dataset.
• Binary categorical features (all numerically-encoded).
• Engineered features from other multi-level categorical features and mixed features.

Make sure that for any new columns that you have engineered, that you've excluded the original columns from the final dataset. Otherwise, their values will interfere with the analysis later on the project. For example, you should not keep "PRAEGENDE_JUGENDJAHRE", since its values won't be useful for the algorithm: only the values derived from it in the engineered features you created should be retained. As a reminder, your data should only be from the subset with few or no missing values.

# If there are other re-engineering tasks you need to perform, make sure you
# take care of them here. (Dealing with missing data will come in step 2.1.)

# droping mixed columns  
df_few_missing.drop(['PRAEGENDE_JUGENDJAHRE', 'CAMEO_INTL_2015',  'LP_LEBENSPHASE_FEIN', 'LP_LEBENSPHASE_GROB'], axis=1, inplace=True)

# Do whatever you need to in order to ensure that the dataframe only contains
# the columns that should be passed to the algorithm functions.

for c in df_few_missing:
    print(c, df_few_missing[c].unique())

ALTERSKATEGORIE_GROB [ 1.  3.  4.  2. nan]
ANREDE_KZ [2 1]
FINANZ_MINIMALIST [1 4 3 2 5]
FINANZ_SPARER [5 4 2 3 1]
FINANZ_VORSORGER [2 1 5 4 3]
FINANZ_ANLEGER [5 2 1 4 3]
FINANZ_UNAUFFAELLIGER [4 3 1 2 5]
FINANZ_HAUSBAUER [5 2 3 4 1]
GREEN_AVANTGARDE [0 1]
HEALTH_TYP [ 3.  2.  1. nan]
RETOURTYP_BK_S [1. 3. 2. 5. 4.]
SEMIO_SOZ [5 4 6 2 7 3 1]
SEMIO_FAM [4 1 5 7 2 6 3]
SEMIO_REL [4 3 2 7 5 1 6]
SEMIO_MAT [3 1 2 4 7 5 6]
SEMIO_VERT [1 4 7 2 6 5 3]
SEMIO_LUST [2 4 6 7 3 1 5]
SEMIO_ERL [2 6 7 4 5 1 3]
SEMIO_KULT [3 4 6 5 7 1 2]
SEMIO_RAT [6 4 3 2 7 5 1]
SEMIO_KRIT [4 7 3 1 5 6 2]
SEMIO_DOM [7 4 2 1 5 6 3]
SEMIO_KAEM [4 7 5 2 3 6 1]
SEMIO_PFLICHT [7 3 4 5 1 6 2]
SEMIO_TRADV [6 3 4 2 7 5 1]
SOHO_KZ [1. 0.]
VERS_TYP [2. 1.]
ANZ_PERSONEN [ 2.  1.  0.  4.  3.  5.  6.  7.  8. 12.  9. 21. 10. 13. 11. 14. 45. 20.
 31. 37. 16. 22. 15. 23. 18. 35. 17. 40. 38.]
ANZ_TITEL [0. 1. 2. 4. 3.]
HH_EINKOMMEN_SCORE [6. 4. 1. 5. 3. 2.]
W_KEIT_KIND_HH [ 3. nan  2.  6.  5.  4.  1.]
WOHNDAUER_2008 [9. 8. 3. 4. 5. 6. 2. 7. 1.]
ANZ_HAUSHALTE_AKTIV [ 11.  10.   1.   3.   5.   4.   6.   2.   9.  nan  14.  13.   8.   7.
  44.  25.  36.  16. 272.  33.  15.  12.  22.  21.  20.  23.  18.  24.
  31.  39.  19.  30.  17.  29.  61.  54.  88.  49.  27.  60.  38.  34.
  48.  35.  62.  26.  45.  42.  28.  77.  47.  50. 236. 137.  43.  52.
  41.  76.  63.  32.  40.  37.  59.  53.  57. 118.  95.  78.  70.  46.
 185.  73.  67. 154.  80. 102.  92. 126.  94.  71.  66.  68.  81.  69.
 121. 258.  58.  86. 124.  85.  90. 128.  51.  72.  64.  56.  83. 150.
 259.  91. 160. 134. 169. 246.  82. 114. 183. 116.  74. 143.  99. 216.
  98. 125. 112. 123.  55. 120.  75. 141. 103.  87.  89. 223.  93. 179.
 135. 149.  97. 101. 110.  79. 111. 109.  65.  84. 367. 162. 195. 107.
 119. 331. 122. 161. 157. 175. 177. 136. 104. 316. 127. 100. 180. 155.
 133. 178. 176. 108. 353. 131. 130.  96. 256. 117. 146. 211. 199. 140.
 200. 106. 285. 263. 253. 163. 377. 171. 148. 191. 243. 153. 145. 174.
 172. 314. 190. 142. 317. 244. 189. 286. 105. 170. 215. 321. 113. 115.
 156. 129. 242. 132. 280. 159. 138. 158. 267. 151. 164. 187. 206. 266.
 265. 139. 193. 209. 184. 210. 260. 366. 194. 318. 344. 277. 595. 197.
 247. 165. 283. 326. 181. 186. 438. 222. 144. 378. 237. 152. 208. 274.
 430. 202. 214. 231. 219. 225. 230. 198. 224. 333. 168. 221. 523. 254.
 305. 166. 252. 445. 147. 301. 213. 536. 220. 249. 290. 515. 276. 241.
 196. 204. 250. 218. 226. 201. 240. 404. 348. 328. 228. 304. 167. 346.
 205. 207. 293. 255. 307. 414. 311. 347. 379. 395. 229. 232.]
ANZ_HH_TITEL [ 0.  1.  5. nan  2.  3.  7.  4.  6.  9. 15. 14.  8. 11. 10. 12. 13. 20.
 16. 17. 23. 18.]
KONSUMNAEHE [ 1.  5.  4.  3.  2.  6.  7. nan]
MIN_GEBAEUDEJAHR [1992. 1997. 2005. 2009. 1994. 1996. 1991. 1993. 1995. 2003. 2006. 2000.
 1990. 2004. 1999. 1998. 2001. 2007. 2002. 1989. 1987. 2008. 1988. 1985.
 2010. 1986.]
OST_WEST_KZ [0 1]
WOHNLAGE [4. 2. 7. 3. 5. 1. 8. 0.]
KBA05_ANTG1 [ 0.  1.  4.  2.  3. nan]
KBA05_ANTG2 [ 0.  3.  1.  4.  2. nan]
KBA05_ANTG3 [ 0.  1. nan  2.  3.]
KBA05_ANTG4 [ 2.  0. nan  1.]
KBA05_GBZ [ 1.  3.  4.  5.  2. nan]
BALLRAUM [ 6.  2.  4.  3.  7.  1.  5. nan]
EWDICHTE [ 3.  4.  2.  5.  6.  1. nan]
INNENSTADT [ 8.  4.  6.  1.  7.  3.  2.  5. nan]
GEBAEUDETYP_RASTER [ 3.  4.  5.  2.  1. nan]
KKK [ 2. nan  3.  4.  1.]
MOBI_REGIO [ 1.  3.  4.  5. nan  2.  6.]
ONLINE_AFFINITAET [3. 2. 1. 5. 4. 0.]
REGIOTYP [ 3.  2. nan  5.  1.  7.  6.  4.]
KBA13_ANZAHL_PKW [963. 712. 596. ...   7.  30.   4.]
PLZ8_ANTG1 [ 2.  3. nan  1.  4.  0.]
PLZ8_ANTG2 [ 3.  2.  4.  1. nan  0.]
PLZ8_ANTG3 [ 2.  1. nan  3.  0.]
PLZ8_ANTG4 [ 1.  0. nan  2.]
PLZ8_BAUMAX [ 1.  2. nan  4.  5.  3.]
PLZ8_HHZ [ 5.  4.  3. nan  2.  1.]
PLZ8_GBZ [ 4.  3.  5. nan  2.  1.]
ARBEIT [ 3.  2.  4.  1.  5. nan]
ORTSGR_KLS9 [ 5.  3.  6.  4.  8.  2.  7.  9.  1. nan]
RELAT_AB [ 4.  2.  3.  5.  1. nan]
CJT_GESAMTTYP_1.0 [0 1]
CJT_GESAMTTYP_2.0 [0 1]
CJT_GESAMTTYP_3.0 [0 1]
CJT_GESAMTTYP_4.0 [0 1]
CJT_GESAMTTYP_5.0 [1 0]
CJT_GESAMTTYP_6.0 [0 1]
FINANZTYP_1 [1 0]
FINANZTYP_2 [0 1]
FINANZTYP_3 [0 1]
FINANZTYP_4 [0 1]
FINANZTYP_5 [0 1]
FINANZTYP_6 [0 1]
GFK_URLAUBERTYP_1.0 [0 1]
GFK_URLAUBERTYP_2.0 [0 1]
GFK_URLAUBERTYP_3.0 [0 1]
GFK_URLAUBERTYP_4.0 [0 1]
GFK_URLAUBERTYP_5.0 [0 1]
GFK_URLAUBERTYP_6.0 [0 1]
GFK_URLAUBERTYP_7.0 [0 1]
GFK_URLAUBERTYP_8.0 [0 1]
GFK_URLAUBERTYP_9.0 [0 1]
GFK_URLAUBERTYP_10.0 [1 0]
GFK_URLAUBERTYP_11.0 [0 1]
GFK_URLAUBERTYP_12.0 [0 1]
LP_FAMILIE_FEIN_1.0 [0 1]
LP_FAMILIE_FEIN_2.0 [0 1]
LP_FAMILIE_FEIN_3.0 [0 1]
LP_FAMILIE_FEIN_4.0 [0 1]
LP_FAMILIE_FEIN_5.0 [1 0]
LP_FAMILIE_FEIN_6.0 [0 1]
LP_FAMILIE_FEIN_7.0 [0 1]
LP_FAMILIE_FEIN_8.0 [0 1]
LP_FAMILIE_FEIN_9.0 [0 1]
LP_FAMILIE_FEIN_10.0 [0 1]
LP_FAMILIE_FEIN_11.0 [0 1]
LP_FAMILIE_GROB_1.0 [0 1]
LP_FAMILIE_GROB_2.0 [0 1]
LP_FAMILIE_GROB_3.0 [1 0]
LP_FAMILIE_GROB_4.0 [0 1]
LP_FAMILIE_GROB_5.0 [0 1]
LP_STATUS_FEIN_1.0 [0 1]
LP_STATUS_FEIN_2.0 [1 0]
LP_STATUS_FEIN_3.0 [0 1]
LP_STATUS_FEIN_4.0 [0 1]
LP_STATUS_FEIN_5.0 [0 1]
LP_STATUS_FEIN_6.0 [0 1]
LP_STATUS_FEIN_7.0 [0 1]
LP_STATUS_FEIN_8.0 [0 1]
LP_STATUS_FEIN_9.0 [0 1]
LP_STATUS_FEIN_10.0 [0 1]
LP_STATUS_GROB_1.0 [1 0]
LP_STATUS_GROB_2.0 [0 1]
LP_STATUS_GROB_3.0 [0 1]
LP_STATUS_GROB_4.0 [0 1]
LP_STATUS_GROB_5.0 [0 1]
NATIONALITAET_KZ_1.0 [1 0]
NATIONALITAET_KZ_2.0 [0 1]
NATIONALITAET_KZ_3.0 [0 1]
SHOPPER_TYP_0.0 [0 1]
SHOPPER_TYP_1.0 [0 1]
SHOPPER_TYP_2.0 [0 1]
SHOPPER_TYP_3.0 [1 0]
ZABEOTYP_1 [0 1]
ZABEOTYP_2 [0 1]
ZABEOTYP_3 [0 1]
ZABEOTYP_4 [0 1]
ZABEOTYP_5 [1 0]
ZABEOTYP_6 [0 1]
GEBAEUDETYP_1.0 [0 1]
GEBAEUDETYP_2.0 [0 1]
GEBAEUDETYP_3.0 [0 1]
GEBAEUDETYP_4.0 [0 1]
GEBAEUDETYP_5.0 [0 1]
GEBAEUDETYP_6.0 [0 1]
GEBAEUDETYP_8.0 [1 0]
CAMEO_DEUG_2015_1 [0 1]
CAMEO_DEUG_2015_2 [0 1]
CAMEO_DEUG_2015_3 [0 1]
CAMEO_DEUG_2015_4 [0 1]
CAMEO_DEUG_2015_5 [0 1]
CAMEO_DEUG_2015_6 [0 1]
CAMEO_DEUG_2015_7 [0 1]
CAMEO_DEUG_2015_8 [1 0]
CAMEO_DEUG_2015_9 [0 1]
CAMEO_DEU_2015_1A [0 1]
CAMEO_DEU_2015_1B [0 1]
CAMEO_DEU_2015_1C [0 1]
CAMEO_DEU_2015_1D [0 1]
CAMEO_DEU_2015_1E [0 1]
CAMEO_DEU_2015_2A [0 1]
CAMEO_DEU_2015_2B [0 1]
CAMEO_DEU_2015_2C [0 1]
CAMEO_DEU_2015_2D [0 1]
CAMEO_DEU_2015_3A [0 1]
CAMEO_DEU_2015_3B [0 1]
CAMEO_DEU_2015_3C [0 1]
CAMEO_DEU_2015_3D [0 1]
CAMEO_DEU_2015_4A [0 1]
CAMEO_DEU_2015_4B [0 1]
CAMEO_DEU_2015_4C [0 1]
CAMEO_DEU_2015_4D [0 1]
CAMEO_DEU_2015_4E [0 1]
CAMEO_DEU_2015_5A [0 1]
CAMEO_DEU_2015_5B [0 1]
CAMEO_DEU_2015_5C [0 1]
CAMEO_DEU_2015_5D [0 1]
CAMEO_DEU_2015_5E [0 1]
CAMEO_DEU_2015_5F [0 1]
CAMEO_DEU_2015_6A [0 1]
CAMEO_DEU_2015_6B [0 1]
CAMEO_DEU_2015_6C [0 1]
CAMEO_DEU_2015_6D [0 1]
CAMEO_DEU_2015_6E [0 1]
CAMEO_DEU_2015_6F [0 1]
CAMEO_DEU_2015_7A [0 1]
CAMEO_DEU_2015_7B [0 1]
CAMEO_DEU_2015_7C [0 1]
CAMEO_DEU_2015_7D [0 1]
CAMEO_DEU_2015_7E [0 1]
CAMEO_DEU_2015_8A [1 0]
CAMEO_DEU_2015_8B [0 1]
CAMEO_DEU_2015_8C [0 1]
CAMEO_DEU_2015_8D [0 1]
CAMEO_DEU_2015_9A [0 1]
CAMEO_DEU_2015_9B [0 1]
CAMEO_DEU_2015_9C [0 1]
CAMEO_DEU_2015_9D [0 1]
CAMEO_DEU_2015_9E [0 1]
PJ_DECADE [ 6.  4.  2.  5.  3. nan  1.]
PJ_MOVEMENT [ 0.  1. nan]
CI_WEALTH [ 5.  2.  1.  4.  3. nan]
CI_LIFESTAGE [ 1.  4.  2.  3.  5. nan]

print(df_few_missing.shape)

(774743, 194)

Step 1.3: Create a Cleaning Function

Even though you've finished cleaning up the general population demographics data, it's important to look ahead to the future and realize that you'll need to perform the same cleaning steps on the customer demographics data. In this substep, complete the function below to execute the main feature selection, encoding, and re-engineering steps you performed above. Then, when it comes to looking at the customer data in Step 3, you can just run this function on that DataFrame to get the trimmed dataset in a single step.

'''
def clean_data(df):
    """
    Perform feature trimming, re-encoding, and engineering for demographics
    data
    
    INPUT: Demographics DataFrame
    OUTPUT: Trimmed and cleaned demographics DataFrame
    """
    
    # Put in code here to execute all main cleaning steps:
    # convert missing value codes into NaNs, ...
    print("1/3  Converting missing value codes into NaNs...")
    for column in df: 
        print(column)
        str_of_unknown = feat_info[feat_info['attribute']==column]['missing_or_unknown'].values[0] 
        list_of_unknown = parser(str_of_unknown) 
    
    for val in list_of_unknown:
        print(val)
        df[column] = df[column].replace(val, np.nan)
        print(df[column].unique())
    
    print("Done.")
    
    
    
    
    # remove selected columns and rows, ...
    print("2/3  Removing selected columns and rows...")

    cols_to_drop = ['AGER_TYP', 'GEBURTSJAHR', 'TITEL_KZ', 'ALTER_HH', 'KK_KUNDENTYP', 'KBA05_BAUMAX']
    df.drop(cols_to_drop, axis=1, inplace=True)
    
    nan_in_rows = df.isnull().sum(axis=1)
    treshold = 9
    rows_below_treshold = df[nan_in_rows <= treshold]
    df_new = rows_below_treshold.copy(deep=True)
    print("Done.")

    
    
    
    # select, re-encode, and engineer column values.
    print("3/3  Selecting, re-encoding, and engineering column values...")

    categorical_cols = feat_info[feat_info['type'] == 'categorical']['attribute'].values
    cat_cols_in_new_df = [x for x in categorical_cols if x in df_new.columns]
    binary = [x for x in cat_cols_in_new_df if df_new[x].nunique()==2]
    multilevel_categorical = [x for x in cat_cols_in_new_df if df_new[x].nunique()>2]
    df_new.loc[:, 'OST_WEST_KZ'].replace({'W':0, 'O':1}, inplace=True)
    num_of_categorical_cols_with_nan = 0
    for col in cat_cols_in_new_df:
        if df_new[col].isnull().values.any() == True: 
            num_of_categorical_cols_with_nan += 1
            df_new[col] = df_new[col].fillna(df_new[col].mode()[0])
    df_new = pd.get_dummies(df_new, columns=multilevel_categorical)
    PRAEGENDE_JUGENDJAHRE_decade = {1: 1 , 2: 1 , 3: 2 ,  4: 2 ,  5: 3 ,  6: 3 ,  7: 3 ,  8: 4 ,  9: 4 ,   10: 5 ,  11: 5 ,  12: 5 , 13: 5 ,  14: 6 , 15: 6 }                             
    PRAEGENDE_JUGENDJAHRE_movement = {1: 0 , 2: 1 ,  3: 0 , 4: 1 , 5: 0 , 6: 1 ,  7: 1 ,  8: 0 ,  9: 1 ,  10: 0 , 11: 1 , 12: 0 , 13: 1 , 14: 0 , 15: 1 }
    CAMEO_INTL_2015_wealth = dict.fromkeys(["11", "12", "13", "14", "15"], 1)
    CAMEO_INTL_2015_wealth.update(dict.fromkeys(["21", "22", "23", "24", "25"], 2))
    CAMEO_INTL_2015_wealth.update(dict.fromkeys(["31", "32", "33", "34", "35"], 3))
    CAMEO_INTL_2015_wealth.update(dict.fromkeys(["41", "42", "43", "44", "45"], 4))
    CAMEO_INTL_2015_wealth.update(dict.fromkeys(["51", "52", "53", "54", "55"], 5))
    CAMEO_INTL_2015_lifeStage = dict.fromkeys(["11", "21", "31", "41", "51"], 1)
    CAMEO_INTL_2015_lifeStage.update(dict.fromkeys(["12", "22", "32", "42", "52"], 2))
    CAMEO_INTL_2015_lifeStage.update(dict.fromkeys(["13", "23", "33", "43", "53"], 3))
    CAMEO_INTL_2015_lifeStage.update(dict.fromkeys(["14", "24", "34", "44", "54"], 4))
    CAMEO_INTL_2015_lifeStage.update(dict.fromkeys(["15", "25", "35", "45", "55"], 5))
    df_new['PJ_DECADE'] = df_new['PRAEGENDE_JUGENDJAHRE']
    df_new['PJ_DECADE'].replace(PRAEGENDE_JUGENDJAHRE_decade, inplace=True)
    df_new['PJ_MOVEMENT'] = df_new['PRAEGENDE_JUGENDJAHRE']
    df_new['PJ_MOVEMENT'].replace(PRAEGENDE_JUGENDJAHRE_movement, inplace=True)
    df_new['CI_WEALTH'] = df_new['CAMEO_INTL_2015']
    df_new['CI_WEALTH'].replace(CAMEO_INTL_2015_wealth, inplace=True)
    df_new['CI_LIFESTAGE'] = df_new['CAMEO_INTL_2015']
    df_new['CI_LIFESTAGE'].replace(CAMEO_INTL_2015_lifeStage, inplace=True)
    df_new.drop(['PRAEGENDE_JUGENDJAHRE', 'CAMEO_INTL_2015',  'LP_LEBENSPHASE_FEIN', 'LP_LEBENSPHASE_GROB'], axis=1, inplace=True)
    print("Done.")


    # Return the cleaned dataframe.
    print("Data cleaned successfully.")

    return df_new
    
'''

def clean_data(df):
    """
    Perform feature trimming, re-encoding, and engineering for demographics
    data
    
    INPUT: Demographics DataFrame
    OUTPUT: Trimmed and cleaned demographics DataFrame
    """
    
    # Put in code here to execute all main cleaning steps:
    # convert missing value codes into NaNs, ...
    print("1/3  Converting missing value codes into NaNs...")
    for column in df: 
        print(column)
        print("Unique before cleaning: ", df[column].unique())
        str_of_unknown = feat_info[feat_info['attribute']==column]['missing_or_unknown'].values[0] 
        list_of_unknown = parser(str_of_unknown) 
    
        for val in list_of_unknown:
            print(val)
            df[column].replace(val, np.nan, inplace=True)
            print(df[column].unique())
    
    print("Done.")
    
    
    
    
    # remove selected columns and rows, ...
    print("2/3  Removing selected columns and rows...")

    cols_to_drop = ['AGER_TYP', 'GEBURTSJAHR', 'TITEL_KZ', 'ALTER_HH', 'KK_KUNDENTYP', 'KBA05_BAUMAX']
    df.drop(cols_to_drop, axis=1, inplace=True)
    
    nan_in_rows = df.isnull().sum(axis=1)
    treshold = 9
    rows_below_treshold = df[nan_in_rows <= treshold]
    df_new = rows_below_treshold.copy(deep=True)
    df = df_new.copy(deep=True)
    print("Done.")

    
    
    
    # select, re-encode, and engineer column values.
    print("3/3  Selecting, re-encoding, and engineering column values...")

    categorical_cols = feat_info[feat_info['type'] == 'categorical']['attribute'].values
    cat_cols_in_new_df = [x for x in categorical_cols if x in df.columns]
    binary = [x for x in cat_cols_in_new_df if df[x].nunique()==2]
    multilevel_categorical = [x for x in cat_cols_in_new_df if df[x].nunique()>2]
    df.loc[:, 'OST_WEST_KZ'].replace({'W':0, 'O':1}, inplace=True)
    num_of_categorical_cols_with_nan = 0
    for col in cat_cols_in_new_df:
        if df[col].isnull().values.any() == True: 
            num_of_categorical_cols_with_nan += 1
            df[col] = df[col].fillna(df[col].mode()[0])
    df = pd.get_dummies(df, columns=multilevel_categorical)
    PRAEGENDE_JUGENDJAHRE_decade = {1: 1 , 2: 1 , 3: 2 ,  4: 2 ,  5: 3 ,  6: 3 ,  7: 3 ,  8: 4 ,  9: 4 ,   10: 5 ,  11: 5 ,  12: 5 , 13: 5 ,  14: 6 , 15: 6 }                             
    PRAEGENDE_JUGENDJAHRE_movement = {1: 0 , 2: 1 ,  3: 0 , 4: 1 , 5: 0 , 6: 1 ,  7: 1 ,  8: 0 ,  9: 1 ,  10: 0 , 11: 1 , 12: 0 , 13: 1 , 14: 0 , 15: 1 }
    CAMEO_INTL_2015_wealth = dict.fromkeys(["11", "12", "13", "14", "15"], 1)
    CAMEO_INTL_2015_wealth.update(dict.fromkeys(["21", "22", "23", "24", "25"], 2))
    CAMEO_INTL_2015_wealth.update(dict.fromkeys(["31", "32", "33", "34", "35"], 3))
    CAMEO_INTL_2015_wealth.update(dict.fromkeys(["41", "42", "43", "44", "45"], 4))
    CAMEO_INTL_2015_wealth.update(dict.fromkeys(["51", "52", "53", "54", "55"], 5))
    CAMEO_INTL_2015_lifeStage = dict.fromkeys(["11", "21", "31", "41", "51"], 1)
    CAMEO_INTL_2015_lifeStage.update(dict.fromkeys(["12", "22", "32", "42", "52"], 2))
    CAMEO_INTL_2015_lifeStage.update(dict.fromkeys(["13", "23", "33", "43", "53"], 3))
    CAMEO_INTL_2015_lifeStage.update(dict.fromkeys(["14", "24", "34", "44", "54"], 4))
    CAMEO_INTL_2015_lifeStage.update(dict.fromkeys(["15", "25", "35", "45", "55"], 5))
    df['PJ_DECADE'] = df['PRAEGENDE_JUGENDJAHRE']
    df['PJ_DECADE'].replace(PRAEGENDE_JUGENDJAHRE_decade, inplace=True)
    df['PJ_MOVEMENT'] = df['PRAEGENDE_JUGENDJAHRE']
    df['PJ_MOVEMENT'].replace(PRAEGENDE_JUGENDJAHRE_movement, inplace=True)
    df['CI_WEALTH'] = df['CAMEO_INTL_2015']
    df['CI_WEALTH'].replace(CAMEO_INTL_2015_wealth, inplace=True)
    df['CI_LIFESTAGE'] = df['CAMEO_INTL_2015']
    df['CI_LIFESTAGE'].replace(CAMEO_INTL_2015_lifeStage, inplace=True)
    df.drop(['PRAEGENDE_JUGENDJAHRE', 'CAMEO_INTL_2015',  'LP_LEBENSPHASE_FEIN', 'LP_LEBENSPHASE_GROB'], axis=1, inplace=True)
    print("Done.")
    

    # Return the cleaned dataframe.
    print("Data cleaned successfully.")

    return df

Step 2: Feature Transformation

Step 2.1: Apply Feature Scaling

Before we apply dimensionality reduction techniques to the data, we need to perform feature scaling so that the principal component vectors are not influenced by the natural differences in scale for features. Starting from this part of the project, you'll want to keep an eye on the API reference page for sklearn to help you navigate to all of the classes and functions that you'll need. In this substep, you'll need to check the following:

• sklearn requires that data not have missing values in order for its estimators to work properly. So, before applying the scaler to your data, make sure that you've cleaned the DataFrame of the remaining missing values. This can be as simple as just removing all data points with missing data, or applying an Imputer to replace all missing values. You might also try a more complicated procedure where you temporarily remove missing values in order to compute the scaling parameters before re-introducing those missing values and applying imputation. Think about how much missing data you have and what possible effects each approach might have on your analysis, and justify your decision in the discussion section below.
• For the actual scaling function, a StandardScaler instance is suggested, scaling each feature to mean 0 and standard deviation 1.
• For these classes, you can make use of the .fit_transform() method to both fit a procedure to the data as well as apply the transformation to the data at the same time. Don't forget to keep the fit sklearn objects handy, since you'll be applying them to the customer demographics data towards the end of the project.

# If you've not yet cleaned the dataset of all NaN values, then investigate and
# do that now.

# CATEGORICAL
cols_filled_mode = feat_info[feat_info['type'] == 'categorical' ]['attribute'].values
mode_cols_in_new_df = [x for x in cols_filled_mode if x in df_few_missing.columns]
for col in mode_cols_in_new_df:
    if df_few_missing[col].isnull().values.any() == True: 
        df_few_missing[col].fillna(df_few_missing[col].mode()[0], inplace=True)
        
# ORDINAL
cols_filled_mode = feat_info[feat_info['type'] == 'ordinal']['attribute'].values
mode_cols_in_new_df = [x for x in cols_filled_mode if x in df_few_missing.columns]
for col in mode_cols_in_new_df:
    if df_few_missing[col].isnull().values.any() == True: 
        df_few_missing[col].fillna(df_few_missing[col].mode()[0], inplace=True)
        
        
# INTERVAL
cols_filled_mode = feat_info[feat_info['type'] == 'interval']['attribute'].values
mode_cols_in_new_df = [x for x in cols_filled_mode if x in df_few_missing.columns]
for col in mode_cols_in_new_df:
    if df_few_missing[col].isnull().values.any() == True: 
        df_few_missing[col].fillna(df_few_missing[col].mode()[0], inplace=True)
        
        
# MIXED
cols_filled_mode = feat_info[feat_info['type'] == 'mixed']['attribute'].values
mode_cols_in_new_df = [x for x in cols_filled_mode if x in df_few_missing.columns]
for col in mode_cols_in_new_df:
    if df_few_missing[col].isnull().values.any() == True: 
        df_few_missing[col].fillna(df_few_missing[col].mode()[0], inplace=True)        

        
# NEW FROM MIXED
mixed = ['PJ_DECADE', 'PJ_MOVEMENT', 'CI_WEALTH', 'CI_LIFESTAGE']
for col in mixed:
    if df_few_missing[col].isnull().values.any() == True: 
        df_few_missing[col].fillna(df_few_missing[col].mode()[0], inplace=True)
        

        
        
# NUMERIC

cols_filled_median = feat_info[feat_info['type'] == 'numeric']['attribute'].values
median_cols_in_new_df = [x for x in cols_filled_median if x in df_few_missing.columns]
for col in median_cols_in_new_df:
    if df_few_missing[col].isnull().values.any() == True: 
        df_few_missing[col].fillna(df_few_missing[col].median(), inplace=True)

'''
def missing_values(df):
    
    # CATEGORICAL
    cols_filled_mode = feat_info[feat_info['type'] == 'categorical' ]['attribute'].values
    mode_cols_in_new_df = [x for x in cols_filled_mode if x in df.columns]
    for col in mode_cols_in_new_df:
        if df[col].isnull().values.any() == True: 
            df[col].fillna(df[col].mode()[0], inplace=True)
            
    # ORDINAL
    cols_filled_mode = feat_info[feat_info['type'] == 'ordinal']['attribute'].values
    mode_cols_in_new_df = [x for x in cols_filled_mode if x in df.columns]
    for col in mode_cols_in_new_df:
        if df[col].isnull().values.any() == True: 
            df[col].fillna(df[col].mode()[0], inplace=True)
            
            
    # INTERVAL
    cols_filled_mode = feat_info[feat_info['type'] == 'interval']['attribute'].values
    mode_cols_in_new_df = [x for x in cols_filled_mode if x in df.columns]
    for col in mode_cols_in_new_df:
        if df[col].isnull().values.any() == True: 
            df[col].fillna(df[col].mode()[0], inplace=True)
            
            
    # MIXED
    cols_filled_mode = feat_info[feat_info['type'] == 'mixed']['attribute'].values
    mode_cols_in_new_df = [x for x in cols_filled_mode if x in df.columns]
    for col in mode_cols_in_new_df:
        if df[col].isnull().values.any() == True: 
            df[col].fillna(df[col].mode()[0], inplace=True)        
    
            
    # NEW FROM MIXED
    mixed = ['PJ_DECADE', 'PJ_MOVEMENT', 'CI_WEALTH', 'CI_LIFESTAGE']
    for col in mixed:
        if df[col].isnull().values.any() == True: 
            df[col].fillna(df[col].mode()[0], inplace=True)
            
            
    # NUMERIC    
    cols_filled_median = feat_info[feat_info['type'] == 'numeric']['attribute'].values
    median_cols_in_new_df = [x for x in cols_filled_median if x in df.columns]
    for col in median_cols_in_new_df:
        if df[col].isnull().values.any() == True: 
            df[col].fillna(df[col].median(), inplace=True)
            
    return df

'''

def missing_values(df):
    
    # CATEGORICAL
    cols_filled_mode = feat_info[feat_info['type'] == 'categorical' ]['attribute'].values
    mode_cols_in_new_df = [x for x in cols_filled_mode if x in df.columns]
    for col in mode_cols_in_new_df:
        if df[col].isnull().values.any() == True: 
            df[col].fillna(df[col].mode()[0], inplace=True)
            
    # ORDINAL
    cols_filled_mode = feat_info[feat_info['type'] == 'ordinal']['attribute'].values
    mode_cols_in_new_df = [x for x in cols_filled_mode if x in df.columns]
    for col in mode_cols_in_new_df:
        if df[col].isnull().values.any() == True: 
            df[col].fillna(df[col].mode()[0], inplace=True)
            
            
    # INTERVAL
    cols_filled_mode = feat_info[feat_info['type'] == 'interval']['attribute'].values
    mode_cols_in_new_df = [x for x in cols_filled_mode if x in df.columns]
    for col in mode_cols_in_new_df:
        if df[col].isnull().values.any() == True: 
            df[col].fillna(df[col].mode()[0], inplace=True)
            
            
    # MIXED
    cols_filled_mode = feat_info[feat_info['type'] == 'mixed']['attribute'].values
    mode_cols_in_new_df = [x for x in cols_filled_mode if x in df.columns]
    for col in mode_cols_in_new_df:
        if df[col].isnull().values.any() == True: 
            df[col].fillna(df[col].mode()[0], inplace=True)        
    
            
    # NEW FROM MIXED
    mixed = ['PJ_DECADE', 'PJ_MOVEMENT', 'CI_WEALTH', 'CI_LIFESTAGE']
    for col in mixed:
        if df[col].isnull().values.any() == True: 
            df[col].fillna(df[col].mode()[0], inplace=True)
            
            
    # NUMERIC    
    cols_filled_median = feat_info[feat_info['type'] == 'numeric']['attribute'].values
    median_cols_in_new_df = [x for x in cols_filled_median if x in df.columns]
    for col in median_cols_in_new_df:
        if df[col].isnull().values.any() == True: 
            df[col].fillna(df[col].median(), inplace=True)
            
    return df

# checking for any null value

for col in df_few_missing:
    if df_few_missing[col].isnull().values.any() == True: 
        print(col, df_few_missing[col].unique())

# Apply feature scaling to the general population demographics data.

from sklearn.preprocessing import StandardScaler

scaler = StandardScaler()
df_std = pd.DataFrame(scaler.fit_transform(df_few_missing), columns=df_few_missing.columns , index = df_few_missing.index)
df_std.head()

	ALTERSKATEGORIE_GROB	ANREDE_KZ	FINANZ_MINIMALIST	FINANZ_SPARER	FINANZ_VORSORGER	FINANZ_ANLEGER	FINANZ_UNAUFFAELLIGER	FINANZ_HAUSBAUER	GREEN_AVANTGARDE	HEALTH_TYP	...	CAMEO_DEU_2015_8D	CAMEO_DEU_2015_9A	CAMEO_DEU_2015_9B	CAMEO_DEU_2015_9C	CAMEO_DEU_2015_9D	CAMEO_DEU_2015_9E	PJ_DECADE	PJ_MOVEMENT	CI_WEALTH	CI_LIFESTAGE
1	-1.765558	0.958422	-1.488785	1.540017	-1.044684	1.468541	0.981043	1.327209	-0.531136	1.007436	...	-0.151294	-0.16386	-0.191328	-0.181489	-0.194409	-0.089107	1.106006	-0.531136	1.170872	-1.249876
2	0.198811	0.958422	-1.488785	0.867244	-1.770775	-0.568811	0.263375	1.327209	1.882757	1.007436	...	-0.151294	-0.16386	-0.191328	-0.181489	-0.194409	-0.089107	1.106006	1.882757	-0.874330	0.763676
3	1.180995	0.958422	0.689233	-0.478302	1.133590	-0.568811	-1.171961	-0.818216	-0.531136	-0.315096	...	-0.151294	-0.16386	-0.191328	-0.181489	-0.194409	-0.089107	-0.259273	-0.531136	-1.556065	-0.578692
4	0.198811	-1.043381	0.689233	0.194471	0.407498	-1.247929	0.263375	-0.818216	-0.531136	1.007436	...	-0.151294	-0.16386	-0.191328	-0.181489	-0.194409	-0.089107	-0.259273	-0.531136	0.489138	0.092492
5	-1.765558	0.958422	-0.036773	-1.151076	1.133590	-0.568811	-0.454293	1.327209	-0.531136	1.007436	...	-0.151294	-0.16386	-0.191328	-0.181489	-0.194409	-0.089107	-1.624552	-0.531136	1.170872	0.763676

5 rows × 194 columns

Discussion 2.1: Apply Feature Scaling

For all columns (except 'numeric') I used the most common value method (filling by mode). It is one of the simpler methods when we want to quickly and manually fill in the gaps (as opposed to machine learning tools, e.g. kNN). For numeric columns, I used median filling to disrupt the distribution of the variables as little as possible.

Scaling changes the feature values so that each column is treated by the model with the same priority. Scaling prevents high-valued features being considered more important, than features, which values are naturaly smaller. After scaling, all features are treated equally.

Step 2.2: Perform Dimensionality Reduction

On your scaled data, you are now ready to apply dimensionality reduction techniques.

• Use sklearn's PCA class to apply principal component analysis on the data, thus finding the vectors of maximal variance in the data. To start, you should not set any parameters (so all components are computed) or set a number of components that is at least half the number of features (so there's enough features to see the general trend in variability).
• Check out the ratio of variance explained by each principal component as well as the cumulative variance explained. Try plotting the cumulative or sequential values using matplotlib's plot() function. Based on what you find, select a value for the number of transformed features you'll retain for the clustering part of the project.
• Once you've made a choice for the number of components to keep, make sure you re-fit a PCA instance to perform the decided-on transformation.

# Apply PCA to the data.

pca = PCA()
result = pca.fit_transform(df_std)

# Investigate the variance accounted for by each principal component.

fig, (ax1, ax2) = plt.subplots(1, 2,figsize=(20,5))
ax1.set_title('Variance explained by each component')
ax1.bar(range(len(pca.explained_variance_ratio_)), pca.explained_variance_ratio_)
ax1.set_xlabel("Principal Component")
ax1.set_ylabel("Variance explained")
ax2.set_title('Cumulative variance explained')
ax2.plot(range(len(pca.explained_variance_ratio_)),np.cumsum(pca.explained_variance_ratio_))
ax2.set_xlabel("Principal Component")
ax2.set_ylabel("Variance explained")

Text(0, 0.5, 'Variance explained')

# Re-apply PCA to the data while selecting for number of components to retain.

pca125 = PCA(n_components=125)
df_pca = pca125.fit_transform(df_std)

Discussion 2.2: Perform Dimensionality Reduction

I chose 125 principal components as they cover more than 90% of the variance.

Step 2.3: Interpret Principal Components

Now that we have our transformed principal components, it's a nice idea to check out the weight of each variable on the first few components to see if they can be interpreted in some fashion.

As a reminder, each principal component is a unit vector that points in the direction of highest variance (after accounting for the variance captured by earlier principal components). The further a weight is from zero, the more the principal component is in the direction of the corresponding feature. If two features have large weights of the same sign (both positive or both negative), then increases in one tend expect to be associated with increases in the other. To contrast, features with different signs can be expected to show a negative correlation: increases in one variable should result in a decrease in the other.

• To investigate the features, you should map each weight to their corresponding feature name, then sort the features according to weight. The most interesting features for each principal component, then, will be those at the beginning and end of the sorted list. Use the data dictionary document to help you understand these most prominent features, their relationships, and what a positive or negative value on the principal component might indicate.
• You should investigate and interpret feature associations from the first three principal components in this substep. To help facilitate this, you should write a function that you can call at any time to print the sorted list of feature weights, for the i-th principal component. This might come in handy in the next step of the project, when you interpret the tendencies of the discovered clusters.

# Map weights for the first principal component to corresponding feature names
# and then print the linked values, sorted by weight.
# HINT: Try defining a function here or in a new cell that you can reuse in the
# other cells.

def pca_weights_mapping(data, pca, i):
    components = pd.DataFrame(pca.components_, columns=list(data.columns))
    weights = components.iloc[i].sort_values(ascending=False)
    return weights

w1 = pca_weights_mapping(df_std, pca125, 0)
print(w1)

LP_STATUS_GROB_1.0    0.193268
PLZ8_ANTG3            0.184002
CI_WEALTH             0.182664
HH_EINKOMMEN_SCORE    0.181283
PLZ8_ANTG4            0.177787
                        ...   
KBA05_GBZ            -0.180126
KBA05_ANTG1          -0.181868
PLZ8_ANTG1           -0.184744
FINANZ_MINIMALIST    -0.185608
MOBI_REGIO           -0.191457
Name: 0, Length: 194, dtype: float64

# Map weights for the second principal component to corresponding feature names
# and then print the linked values, sorted by weight.

w2 = pca_weights_mapping(df_std, pca125, 1)
print(w2)

ALTERSKATEGORIE_GROB     0.234395
FINANZ_VORSORGER         0.219813
ZABEOTYP_3               0.199029
SEMIO_ERL                0.180901
SEMIO_LUST               0.163867
                           ...   
SEMIO_TRADV             -0.208826
FINANZ_UNAUFFAELLIGER   -0.215114
SEMIO_REL               -0.216490
FINANZ_SPARER           -0.229111
PJ_DECADE               -0.232007
Name: 1, Length: 194, dtype: float64

# Map weights for the third principal component to corresponding feature names
# and then print the linked values, sorted by weight.

w3 = pca_weights_mapping(df_std, pca125, 2)
print(w3)

SEMIO_VERT     0.315931
SEMIO_FAM      0.256153
SEMIO_SOZ      0.254743
SEMIO_KULT     0.246449
FINANZTYP_5    0.138459
                 ...   
SEMIO_ERL     -0.205590
SEMIO_KRIT    -0.263509
SEMIO_DOM     -0.275249
SEMIO_KAEM    -0.308958
ANREDE_KZ     -0.339670
Name: 2, Length: 194, dtype: float64

Discussion 2.3: Interpret Principal Components

At the outset, I must point out that the weights obtained indicate that no Principal Component is strongly correlated with any feature from the original data set. The highest absolute value of the correlation is only 0.3397 for ANREDE_KZ and 3rd Principal Component and it is not a high value.

(PCA1) First Principal Component Analysis

Description of the first and last 3 features:

LP_STATUS_GROB_1.0 - social status (the more - the more wealthy)

PLZ8_ANTG3 - number of 6-10 family houses in the region (the more - the higher number)

CI_WEALTH - level of how wealthy is a household (increase - poorer)

(4th feature, for making sure) HH_EINKOMMEN_SCORE - estimated household net income (the more - the lower income)

PLZ8_ANTG1 - number of 1-2 family houses in the region (the more - the higher number)

FINANZ_MINIMALIST - low financial interest (the more points - the more true)

MOBI_REGIO - movement patterns (from highiest to lowest)

On the basis of the first three and the last three features, I conclude that this component is a measure of the financial status, financial awareness and environmental factors in the region. However, I am somewhat consterned by the fact that two features: LP_STATUS_GROB_1.0 and CI_WEALTH are mutually correlated, although the description of their value contradicts this (the higher the value of LP, the better the material position AND the higher the value of CI, the worse the material position). The explanation for this may be a wrong description of the LP scale or the fact that one of them may apply to a single person and the other - to the entire household. The next, fourth feature (HH_EINKOMMEN_SCORE) confirms the CI_WEALTH feature, so I will consider it as the correct correlation.

Generally:

smaller PCA1 value =

• more 1- or 2-family houses (I assume this describes the peaceful, quiet neighborhood of small family houses)
• less interest in finances
• less movement
• richer household

greater PCA1 =

• more 6-10 family houses (I assume they are blocks of flats, tenement houses, smaller urban estates)
• no data showing greater interest in finances
• greater movement
• poorer household

---

(PCA2) Second Principal Component Analysis

Description of the first and last 3 features:

ALTERSKATEGORIE_GROB - Estimated age based on given name analysis (the more - the older)

FINANZ_VORSORGER - stable money status, has money, but not investor yet (the more points - the more true)

ZABEOTYP_3 - Energy consumption typology (the less - the more "green", the more - the more indifferent to the topic)

SEMIO_REL - religious (the more points - the less affinity)

FINANZ_SPARER - money-saver (the more points - the more true)

PJ_DECADE - dominating decade of person's youth (the more - the later years)

The Second Principal Component is a measure of a person's age, the way they manage their money and the values they follow (environmental protection, economy, religiosity).

Generally:

smaller PCA1 value =

• younger person
• greener, focuses on energy consumption
• less religious
• big link with saving money
• later decades of youth (obviously, since younger people)

greater PCA1 =

• a more elderly person
• more "FINANZ_VORSORGER", that is, more stable money status
• more indifferent to energy consumption
• more religious
• less to do with radical money saving
• earlier decades of youth (obviously, since older people)

---

(PCA3) Third Principal Component Analysis

Description of the first and last 3 features:

SEMIO_VERT - dreamful (the more points - the less affinity)

SEMIO_FAM - family-minded (the more points - the less affinity)

SEMIO_SOZ - socially-minded (the more points - the less affinity)

SEMIO_DOM - dominant-minded (the more points - the less affinity)

SEMIO_KAEM - combative attitude (the more points - the less affinity)

ANREDE_KZ - gender (male 1, female 2)

Third Principal Component is a measure of personality typology.

Generally:

smaller PCA1 value =

• greater probability that the data relates to a woman
• more dreamful
• more family-minded
• more socially-minded
• less dominant-minded
• less combative attitude

greater PCA1 =

• greater probability that the data relates to a man
• less dreamful
• less family-minded
• less socially-minded
• more dominant-minded
• more combative attitude

Step 3: Clustering

Step 3.1: Apply Clustering to General Population

You've assessed and cleaned the demographics data, then scaled and transformed them. Now, it's time to see how the data clusters in the principal components space. In this substep, you will apply k-means clustering to the dataset and use the average within-cluster distances from each point to their assigned cluster's centroid to decide on a number of clusters to keep.

• Use sklearn's KMeans class to perform k-means clustering on the PCA-transformed data.
• Then, compute the average difference from each point to its assigned cluster's center. Hint: The KMeans object's .score() method might be useful here, but note that in sklearn, scores tend to be defined so that larger is better. Try applying it to a small, toy dataset, or use an internet search to help your understanding.
• Perform the above two steps for a number of different cluster counts. You can then see how the average distance decreases with an increasing number of clusters. However, each additional cluster provides a smaller net benefit. Use this fact to select a final number of clusters in which to group the data. Warning: because of the large size of the dataset, it can take a long time for the algorithm to resolve. The more clusters to fit, the longer the algorithm will take. You should test for cluster counts through at least 10 clusters to get the full picture, but you shouldn't need to test for a number of clusters above about 30.
• Once you've selected a final number of clusters to use, re-fit a KMeans instance to perform the clustering operation. Make sure that you also obtain the cluster assignments for the general demographics data, since you'll be using them in the final Step 3.3.

# Over a number of different cluster counts...
    # run k-means clustering on the data and...    
    # compute the average within-cluster distances.
    
    
def k_mean_score(data, n):
    kmeans = MiniBatchKMeans(n_clusters=n, batch_size=2048) # mini batch for fast computing, because number of data is much above 10k rows
    model = kmeans.fit(data)
    score = np.abs(model.score(data))
    return score


clusters = 30
scores = []
num_of_clusters = list(range(1, clusters + 1))
for k in num_of_clusters:
    print("Computing {} of {}".format(k, clusters))
    scores.append(k_mean_score(df_pca, k))
    

Computing 1 of 30
Computing 2 of 30
Computing 3 of 30
Computing 4 of 30
Computing 5 of 30
Computing 6 of 30
Computing 7 of 30
Computing 8 of 30
Computing 9 of 30
Computing 10 of 30
Computing 11 of 30
Computing 12 of 30
Computing 13 of 30
Computing 14 of 30
Computing 15 of 30
Computing 16 of 30
Computing 17 of 30
Computing 18 of 30
Computing 19 of 30
Computing 20 of 30
Computing 21 of 30
Computing 22 of 30
Computing 23 of 30
Computing 24 of 30
Computing 25 of 30
Computing 26 of 30
Computing 27 of 30
Computing 28 of 30
Computing 29 of 30
Computing 30 of 30

# Investigate the change in within-cluster distance across number of clusters.
# HINT: Use matplotlib's plot function to visualize this relationship.

fig, ax1 = plt.subplots(1, 1, figsize=(10,5))
ax1.plot(num_of_clusters, scores, marker='o')
ax1.set_title("KMeans")
ax1.set_xlabel('Number of clusters')
ax1.set_ylabel('Sum of Squared Error (SSE)')

Text(0, 0.5, 'Sum of Squared Error (SSE)')

# Re-fit the k-means model with the selected number of clusters and obtain
# cluster predictions for the general population demographics data.

kmeans = MiniBatchKMeans(n_clusters=12, batch_size=2048)
model_12 = kmeans.fit(df_pca)
population_prediction = model_12.predict(df_pca)

Discussion 3.1: Apply Clustering to General Population

The plot from MiniBatch is less smooth than the plot from regular KMeans, because computations and results arre obtained in the series of batches, which makes the single error results more unstable. However, calculations are performed much much (!) faster, which is a significant advantage.

At the plot I can see no clear point for so-called 'elbow'. I subjectively selected point 12th as the one to which the error drops significantly. So I decided to segment the population into 12 clusters on the basis of the scree plot above. Behind that point, the error drops slower. I think that chosing 12 clusters is the most accommodating decision in this case.

Step 3.2: Apply All Steps to the Customer Data

Now that you have clusters and cluster centers for the general population, it's time to see how the customer data maps on to those clusters. Take care to not confuse this for re-fitting all of the models to the customer data. Instead, you're going to use the fits from the general population to clean, transform, and cluster the customer data. In the last step of the project, you will interpret how the general population fits apply to the customer data.

• Don't forget when loading in the customers data, that it is semicolon (;) delimited.
• Apply the same feature wrangling, selection, and engineering steps to the customer demographics using the clean_data() function you created earlier. (You can assume that the customer demographics data has similar meaning behind missing data patterns as the general demographics data.)
• Use the sklearn objects from the general demographics data, and apply their transformations to the customers data. That is, you should not be using a .fit() or .fit_transform() method to re-fit the old objects, nor should you be creating new sklearn objects! Carry the data through the feature scaling, PCA, and clustering steps, obtaining cluster assignments for all of the data in the customer demographics data.

# Load in the customer demographics data.
customers = pd.read_csv('Udacity_CUSTOMERS_Subset.csv',sep=';')

'''
# Identify missing or unknown data values and convert them to NaNs.

def parser(string):
    list_of_values = []
    
    if string == "[]":
        return list_of_values

    string = string.strip(' ').strip('')
    string_without_brackets = string.lstrip("[").rstrip("]")
    list_of_string_values = string_without_brackets.split(",")
    
    print("List: {}; Len {}".format(list_of_string_values, len(list_of_string_values)))

    for i in list_of_string_values:
        if i == 'X' or i == 'XX':
            list_of_values.append(str(i))
        else:
            list_of_values.append(int(i))
        
    return list_of_values

'''

'\n# Identify missing or unknown data values and convert them to NaNs.\n\ndef parser(string):\n    list_of_values = []\n    \n    if string == "[]":\n        return list_of_values\n\n    string = string.strip(\' \').strip(\'\')\n    string_without_brackets = string.lstrip("[").rstrip("]")\n    list_of_string_values = string_without_brackets.split(",")\n    \n    print("List: {}; Len {}".format(list_of_string_values, len(list_of_string_values)))\n\n    for i in list_of_string_values:\n        if i == \'X\' or i == \'XX\':\n            list_of_values.append(str(i))\n        else:\n            list_of_values.append(int(i))\n        \n    return list_of_values\n\n'

'''
def clean_data(df):
    """
    Perform feature trimming, re-encoding, and engineering for demographics
    data
    
    INPUT: Demographics DataFrame
    OUTPUT: Trimmed and cleaned demographics DataFrame
    """
    
    # Put in code here to execute all main cleaning steps:
    # convert missing value codes into NaNs, ...
    print("1/3  Converting missing value codes into NaNs...")
    for column in df: 
        print(column)
        print("Unique before cleaning: ", df[column].unique())
        str_of_unknown = feat_info[feat_info['attribute']==column]['missing_or_unknown'].values[0] 
        list_of_unknown = parser(str_of_unknown) 
    
        for val in list_of_unknown:
            print(val)
            df[column].replace(val, np.nan, inplace=True)
            print(df[column].unique())
    
    print("Done.")
    
    
    
    
    # remove selected columns and rows, ...
    print("2/3  Removing selected columns and rows...")

    cols_to_drop = ['AGER_TYP', 'GEBURTSJAHR', 'TITEL_KZ', 'ALTER_HH', 'KK_KUNDENTYP', 'KBA05_BAUMAX']
    df.drop(cols_to_drop, axis=1, inplace=True)
    
    nan_in_rows = df.isnull().sum(axis=1)
    treshold = 9
    rows_below_treshold = df[nan_in_rows <= treshold]
    df_new = rows_below_treshold.copy(deep=True)
    df = df_new.copy(deep=True)
    print("Done.")

    
    
    
    # select, re-encode, and engineer column values.
    print("3/3  Selecting, re-encoding, and engineering column values...")

    categorical_cols = feat_info[feat_info['type'] == 'categorical']['attribute'].values
    cat_cols_in_new_df = [x for x in categorical_cols if x in df.columns]
    binary = [x for x in cat_cols_in_new_df if df[x].nunique()==2]
    multilevel_categorical = [x for x in cat_cols_in_new_df if df[x].nunique()>2]
    df.loc[:, 'OST_WEST_KZ'].replace({'W':0, 'O':1}, inplace=True)
    num_of_categorical_cols_with_nan = 0
    for col in cat_cols_in_new_df:
        if df[col].isnull().values.any() == True: 
            num_of_categorical_cols_with_nan += 1
            df[col] = df[col].fillna(df[col].mode()[0])
    df = pd.get_dummies(df, columns=multilevel_categorical)
    PRAEGENDE_JUGENDJAHRE_decade = {1: 1 , 2: 1 , 3: 2 ,  4: 2 ,  5: 3 ,  6: 3 ,  7: 3 ,  8: 4 ,  9: 4 ,   10: 5 ,  11: 5 ,  12: 5 , 13: 5 ,  14: 6 , 15: 6 }                             
    PRAEGENDE_JUGENDJAHRE_movement = {1: 0 , 2: 1 ,  3: 0 , 4: 1 , 5: 0 , 6: 1 ,  7: 1 ,  8: 0 ,  9: 1 ,  10: 0 , 11: 1 , 12: 0 , 13: 1 , 14: 0 , 15: 1 }
    CAMEO_INTL_2015_wealth = dict.fromkeys(["11", "12", "13", "14", "15"], 1)
    CAMEO_INTL_2015_wealth.update(dict.fromkeys(["21", "22", "23", "24", "25"], 2))
    CAMEO_INTL_2015_wealth.update(dict.fromkeys(["31", "32", "33", "34", "35"], 3))
    CAMEO_INTL_2015_wealth.update(dict.fromkeys(["41", "42", "43", "44", "45"], 4))
    CAMEO_INTL_2015_wealth.update(dict.fromkeys(["51", "52", "53", "54", "55"], 5))
    CAMEO_INTL_2015_lifeStage = dict.fromkeys(["11", "21", "31", "41", "51"], 1)
    CAMEO_INTL_2015_lifeStage.update(dict.fromkeys(["12", "22", "32", "42", "52"], 2))
    CAMEO_INTL_2015_lifeStage.update(dict.fromkeys(["13", "23", "33", "43", "53"], 3))
    CAMEO_INTL_2015_lifeStage.update(dict.fromkeys(["14", "24", "34", "44", "54"], 4))
    CAMEO_INTL_2015_lifeStage.update(dict.fromkeys(["15", "25", "35", "45", "55"], 5))
    df['PJ_DECADE'] = df['PRAEGENDE_JUGENDJAHRE']
    df['PJ_DECADE'].replace(PRAEGENDE_JUGENDJAHRE_decade, inplace=True)
    df['PJ_MOVEMENT'] = df['PRAEGENDE_JUGENDJAHRE']
    df['PJ_MOVEMENT'].replace(PRAEGENDE_JUGENDJAHRE_movement, inplace=True)
    df['CI_WEALTH'] = df['CAMEO_INTL_2015']
    df['CI_WEALTH'].replace(CAMEO_INTL_2015_wealth, inplace=True)
    df['CI_LIFESTAGE'] = df['CAMEO_INTL_2015']
    df['CI_LIFESTAGE'].replace(CAMEO_INTL_2015_lifeStage, inplace=True)
    df.drop(['PRAEGENDE_JUGENDJAHRE', 'CAMEO_INTL_2015',  'LP_LEBENSPHASE_FEIN', 'LP_LEBENSPHASE_GROB'], axis=1, inplace=True)
    print("Done.")
    

    # Return the cleaned dataframe.
    print("Data cleaned successfully.")

    return df
'''

'\ndef clean_data(df):\n    """\n    Perform feature trimming, re-encoding, and engineering for demographics\n    data\n    \n    INPUT: Demographics DataFrame\n    OUTPUT: Trimmed and cleaned demographics DataFrame\n    """\n    \n    # Put in code here to execute all main cleaning steps:\n    # convert missing value codes into NaNs, ...\n    print("1/3  Converting missing value codes into NaNs...")\n    for column in df: \n        print(column)\n        print("Unique before cleaning: ", df[column].unique())\n        str_of_unknown = feat_info[feat_info[\'attribute\']==column][\'missing_or_unknown\'].values[0] \n        list_of_unknown = parser(str_of_unknown) \n    \n        for val in list_of_unknown:\n            print(val)\n            df[column].replace(val, np.nan, inplace=True)\n            print(df[column].unique())\n    \n    print("Done.")\n    \n    \n    \n    \n    # remove selected columns and rows, ...\n    print("2/3  Removing selected columns and rows...")\n\n    cols_to_drop = [\'AGER_TYP\', \'GEBURTSJAHR\', \'TITEL_KZ\', \'ALTER_HH\', \'KK_KUNDENTYP\', \'KBA05_BAUMAX\']\n    df.drop(cols_to_drop, axis=1, inplace=True)\n    \n    nan_in_rows = df.isnull().sum(axis=1)\n    treshold = 9\n    rows_below_treshold = df[nan_in_rows <= treshold]\n    df_new = rows_below_treshold.copy(deep=True)\n    df = df_new.copy(deep=True)\n    print("Done.")\n\n    \n    \n    \n    # select, re-encode, and engineer column values.\n    print("3/3  Selecting, re-encoding, and engineering column values...")\n\n    categorical_cols = feat_info[feat_info[\'type\'] == \'categorical\'][\'attribute\'].values\n    cat_cols_in_new_df = [x for x in categorical_cols if x in df.columns]\n    binary = [x for x in cat_cols_in_new_df if df[x].nunique()==2]\n    multilevel_categorical = [x for x in cat_cols_in_new_df if df[x].nunique()>2]\n    df.loc[:, \'OST_WEST_KZ\'].replace({\'W\':0, \'O\':1}, inplace=True)\n    num_of_categorical_cols_with_nan = 0\n    for col in cat_cols_in_new_df:\n        if df[col].isnull().values.any() == True: \n            num_of_categorical_cols_with_nan += 1\n            df[col] = df[col].fillna(df[col].mode()[0])\n    df = pd.get_dummies(df, columns=multilevel_categorical)\n    PRAEGENDE_JUGENDJAHRE_decade = {1: 1 , 2: 1 , 3: 2 ,  4: 2 ,  5: 3 ,  6: 3 ,  7: 3 ,  8: 4 ,  9: 4 ,   10: 5 ,  11: 5 ,  12: 5 , 13: 5 ,  14: 6 , 15: 6 }                             \n    PRAEGENDE_JUGENDJAHRE_movement = {1: 0 , 2: 1 ,  3: 0 , 4: 1 , 5: 0 , 6: 1 ,  7: 1 ,  8: 0 ,  9: 1 ,  10: 0 , 11: 1 , 12: 0 , 13: 1 , 14: 0 , 15: 1 }\n    CAMEO_INTL_2015_wealth = dict.fromkeys(["11", "12", "13", "14", "15"], 1)\n    CAMEO_INTL_2015_wealth.update(dict.fromkeys(["21", "22", "23", "24", "25"], 2))\n    CAMEO_INTL_2015_wealth.update(dict.fromkeys(["31", "32", "33", "34", "35"], 3))\n    CAMEO_INTL_2015_wealth.update(dict.fromkeys(["41", "42", "43", "44", "45"], 4))\n    CAMEO_INTL_2015_wealth.update(dict.fromkeys(["51", "52", "53", "54", "55"], 5))\n    CAMEO_INTL_2015_lifeStage = dict.fromkeys(["11", "21", "31", "41", "51"], 1)\n    CAMEO_INTL_2015_lifeStage.update(dict.fromkeys(["12", "22", "32", "42", "52"], 2))\n    CAMEO_INTL_2015_lifeStage.update(dict.fromkeys(["13", "23", "33", "43", "53"], 3))\n    CAMEO_INTL_2015_lifeStage.update(dict.fromkeys(["14", "24", "34", "44", "54"], 4))\n    CAMEO_INTL_2015_lifeStage.update(dict.fromkeys(["15", "25", "35", "45", "55"], 5))\n    df[\'PJ_DECADE\'] = df[\'PRAEGENDE_JUGENDJAHRE\']\n    df[\'PJ_DECADE\'].replace(PRAEGENDE_JUGENDJAHRE_decade, inplace=True)\n    df[\'PJ_MOVEMENT\'] = df[\'PRAEGENDE_JUGENDJAHRE\']\n    df[\'PJ_MOVEMENT\'].replace(PRAEGENDE_JUGENDJAHRE_movement, inplace=True)\n    df[\'CI_WEALTH\'] = df[\'CAMEO_INTL_2015\']\n    df[\'CI_WEALTH\'].replace(CAMEO_INTL_2015_wealth, inplace=True)\n    df[\'CI_LIFESTAGE\'] = df[\'CAMEO_INTL_2015\']\n    df[\'CI_LIFESTAGE\'].replace(CAMEO_INTL_2015_lifeStage, inplace=True)\n    df.drop([\'PRAEGENDE_JUGENDJAHRE\', \'CAMEO_INTL_2015\',  \'LP_LEBENSPHASE_FEIN\', \'LP_LEBENSPHASE_GROB\'], axis=1, inplace=True)\n    print("Done.")\n    \n\n    # Return the cleaned dataframe.\n    print("Data cleaned successfully.")\n\n    return df\n'

'''
def missing_values(df):
    
    # CATEGORICAL
    cols_filled_mode = feat_info[feat_info['type'] == 'categorical' ]['attribute'].values
    mode_cols_in_new_df = [x for x in cols_filled_mode if x in df.columns]
    for col in mode_cols_in_new_df:
        if df[col].isnull().values.any() == True: 
            df[col].fillna(df[col].mode()[0], inplace=True)
            
    # ORDINAL
    cols_filled_mode = feat_info[feat_info['type'] == 'ordinal']['attribute'].values
    mode_cols_in_new_df = [x for x in cols_filled_mode if x in df.columns]
    for col in mode_cols_in_new_df:
        if df[col].isnull().values.any() == True: 
            df[col].fillna(df[col].mode()[0], inplace=True)
            
            
    # INTERVAL
    cols_filled_mode = feat_info[feat_info['type'] == 'interval']['attribute'].values
    mode_cols_in_new_df = [x for x in cols_filled_mode if x in df.columns]
    for col in mode_cols_in_new_df:
        if df[col].isnull().values.any() == True: 
            df[col].fillna(df[col].mode()[0], inplace=True)
            
            
    # MIXED
    cols_filled_mode = feat_info[feat_info['type'] == 'mixed']['attribute'].values
    mode_cols_in_new_df = [x for x in cols_filled_mode if x in df.columns]
    for col in mode_cols_in_new_df:
        if df[col].isnull().values.any() == True: 
            df[col].fillna(df[col].mode()[0], inplace=True)        
    
            
    # NEW FROM MIXED
    mixed = ['PJ_DECADE', 'PJ_MOVEMENT', 'CI_WEALTH', 'CI_LIFESTAGE']
    for col in mixed:
        if df[col].isnull().values.any() == True: 
            df[col].fillna(df[col].mode()[0], inplace=True)
            
            
    # NUMERIC    
    cols_filled_median = feat_info[feat_info['type'] == 'numeric']['attribute'].values
    median_cols_in_new_df = [x for x in cols_filled_median if x in df.columns]
    for col in median_cols_in_new_df:
        if df[col].isnull().values.any() == True: 
            df[col].fillna(df[col].median(), inplace=True)
            
    return df
'''

"\ndef missing_values(df):\n    \n    # CATEGORICAL\n    cols_filled_mode = feat_info[feat_info['type'] == 'categorical' ]['attribute'].values\n    mode_cols_in_new_df = [x for x in cols_filled_mode if x in df.columns]\n    for col in mode_cols_in_new_df:\n        if df[col].isnull().values.any() == True: \n            df[col].fillna(df[col].mode()[0], inplace=True)\n            \n    # ORDINAL\n    cols_filled_mode = feat_info[feat_info['type'] == 'ordinal']['attribute'].values\n    mode_cols_in_new_df = [x for x in cols_filled_mode if x in df.columns]\n    for col in mode_cols_in_new_df:\n        if df[col].isnull().values.any() == True: \n            df[col].fillna(df[col].mode()[0], inplace=True)\n            \n            \n    # INTERVAL\n    cols_filled_mode = feat_info[feat_info['type'] == 'interval']['attribute'].values\n    mode_cols_in_new_df = [x for x in cols_filled_mode if x in df.columns]\n    for col in mode_cols_in_new_df:\n        if df[col].isnull().values.any() == True: \n            df[col].fillna(df[col].mode()[0], inplace=True)\n            \n            \n    # MIXED\n    cols_filled_mode = feat_info[feat_info['type'] == 'mixed']['attribute'].values\n    mode_cols_in_new_df = [x for x in cols_filled_mode if x in df.columns]\n    for col in mode_cols_in_new_df:\n        if df[col].isnull().values.any() == True: \n            df[col].fillna(df[col].mode()[0], inplace=True)        \n    \n            \n    # NEW FROM MIXED\n    mixed = ['PJ_DECADE', 'PJ_MOVEMENT', 'CI_WEALTH', 'CI_LIFESTAGE']\n    for col in mixed:\n        if df[col].isnull().values.any() == True: \n            df[col].fillna(df[col].mode()[0], inplace=True)\n            \n            \n    # NUMERIC    \n    cols_filled_median = feat_info[feat_info['type'] == 'numeric']['attribute'].values\n    median_cols_in_new_df = [x for x in cols_filled_median if x in df.columns]\n    for col in median_cols_in_new_df:\n        if df[col].isnull().values.any() == True: \n            df[col].fillna(df[col].median(), inplace=True)\n            \n    return df\n"

# Apply preprocessing, feature transformation, and clustering from the general
# demographics onto the customer data, obtaining cluster predictions for the
# customer demographics data.

# data cleaning
print("=========================================== Data cleaning ================================================")
df_customers = clean_data(customers)
print("========================================== /Data cleaning ================================================")

print("\n\n")

df_customers.insert(134, 'GEBAEUDETYP_5.0', 0) # that value wasn't present in the customer dataset, so the dummy column
# couldn't appear

#checking if the shape is the same
print(df_customers.shape)
print(df_few_missing.shape)
    

print("Missing values")
# missing values
df_customers = missing_values(df_customers)

# checking for any null value
for col in df_customers:
    if df_customers[col].isnull().values.any() == True: 
        print(col, df_customers[col].unique())
    else:
        print(".")
print("\n\n")


print("Scaling")
# scaling
df_customers_std = pd.DataFrame(scaler.transform(df_customers), 
                                columns=df_customers.columns.tolist(), index=df_customers.index) # without new fitting


print("PCA")
# pca
df_customers_pca = pca125.transform(df_customers_std) # without new fitting


print("K-means")
# k-means
customers_prediction = model_12.predict(df_customers_pca) # without new fitting

=========================================== Data cleaning ================================================
1/3  Converting missing value codes into NaNs...
AGER_TYP
Unique before cleaning:  [ 2 -1  1  0  3]
List: ['-1', '0']; Len 2
-1
[ 2. nan  1.  0.  3.]
0
[ 2. nan  1.  3.]
ALTERSKATEGORIE_GROB
Unique before cleaning:  [4 3 2 1 9]
List: ['-1', '0', '9']; Len 3
-1
[4 3 2 1 9]
0
[4 3 2 1 9]
9
[ 4.  3.  2.  1. nan]
ANREDE_KZ
Unique before cleaning:  [1 2]
List: ['-1', '0']; Len 2
-1
[1 2]
0
[1 2]
CJT_GESAMTTYP
Unique before cleaning:  [ 5. nan  2.  6.  4.  1.  3.]
List: ['0']; Len 1
0
[ 5. nan  2.  6.  4.  1.  3.]
FINANZ_MINIMALIST
Unique before cleaning:  [5 3 2 4 1]
List: ['-1']; Len 1
-1
[5 3 2 4 1]
FINANZ_SPARER
Unique before cleaning:  [1 2 3 4 5]
List: ['-1']; Len 1
-1
[1 2 3 4 5]
FINANZ_VORSORGER
Unique before cleaning:  [5 4 1 2 3]
List: ['-1']; Len 1
-1
[5 4 1 2 3]
FINANZ_ANLEGER
Unique before cleaning:  [1 2 4 3 5]
List: ['-1']; Len 1
-1
[1 2 4 3 5]
FINANZ_UNAUFFAELLIGER
Unique before cleaning:  [2 3 4 1 5]
List: ['-1']; Len 1
-1
[2 3 4 1 5]
FINANZ_HAUSBAUER
Unique before cleaning:  [2 4 3 5 1]
List: ['-1']; Len 1
-1
[2 4 3 5 1]
FINANZTYP
Unique before cleaning:  [2 6 5 3 1 4]
List: ['-1']; Len 1
-1
[2 6 5 3 1 4]
GEBURTSJAHR
Unique before cleaning:  [   0 1960 1942 1938 1914 1959 1943 1964 1961 1923 1953 1985 1963 1965
 1947 1930 1962 1967 1935 1948 1955 1931 1956 1945 1937 1932 1971 1976
 1974 1925 1939 1950 1980 1984 1949 1941 1970 1966 1926 1929 1957 1954
 1934 1977 1995 1936 1946 1928 1958 1979 1940 1933 1951 1972 1952 1968
 1981 1975 1969 1982 1973 1927 1944 1924 1978 1996 2017 1992 1987 1921
 2014 1986 1919 1990 2012 2016 1902 1994 2002 1999 1983 1916 1920 2013
 1991 1922 1918 1989 1988 1997 1993 2015 2000 1998 1915 2003 1910 2001
 1917 1900 2005 2009 2008 1912 2006 1913 1911 2011 2004 2010 2007 1909
 1908]
List: ['0']; Len 1
0
[  nan 1960. 1942. 1938. 1914. 1959. 1943. 1964. 1961. 1923. 1953. 1985.
 1963. 1965. 1947. 1930. 1962. 1967. 1935. 1948. 1955. 1931. 1956. 1945.
 1937. 1932. 1971. 1976. 1974. 1925. 1939. 1950. 1980. 1984. 1949. 1941.
 1970. 1966. 1926. 1929. 1957. 1954. 1934. 1977. 1995. 1936. 1946. 1928.
 1958. 1979. 1940. 1933. 1951. 1972. 1952. 1968. 1981. 1975. 1969. 1982.
 1973. 1927. 1944. 1924. 1978. 1996. 2017. 1992. 1987. 1921. 2014. 1986.
 1919. 1990. 2012. 2016. 1902. 1994. 2002. 1999. 1983. 1916. 1920. 2013.
 1991. 1922. 1918. 1989. 1988. 1997. 1993. 2015. 2000. 1998. 1915. 2003.
 1910. 2001. 1917. 1900. 2005. 2009. 2008. 1912. 2006. 1913. 1911. 2011.
 2004. 2010. 2007. 1909. 1908.]
GFK_URLAUBERTYP
Unique before cleaning:  [ 4. nan  3. 10.  2. 11.  8.  1.  5.  9. 12.  7.  6.]
GREEN_AVANTGARDE
Unique before cleaning:  [1 0]
HEALTH_TYP
Unique before cleaning:  [ 1  2  3 -1]
List: ['-1', '0']; Len 2
-1
[ 1.  2.  3. nan]
0
[ 1.  2.  3. nan]
LP_LEBENSPHASE_FEIN
Unique before cleaning:  [20. nan 13.  0. 31. 17.  6. 28.  5. 27. 40. 35.  2. 19. 38. 36.  8. 34.
 10. 12. 26. 11.  9. 37. 14. 39. 32.  7. 15. 23. 21. 25. 33. 16. 24. 30.
 18.  4. 22.  1. 29.  3.]
List: ['0']; Len 1
0
[20. nan 13. 31. 17.  6. 28.  5. 27. 40. 35.  2. 19. 38. 36.  8. 34. 10.
 12. 26. 11.  9. 37. 14. 39. 32.  7. 15. 23. 21. 25. 33. 16. 24. 30. 18.
  4. 22.  1. 29.  3.]
LP_LEBENSPHASE_GROB
Unique before cleaning:  [ 5. nan  3.  0. 10.  2.  8. 12. 11.  1.  4.  6.  7.  9.]
List: ['0']; Len 1
0
[ 5. nan  3. 10.  2.  8. 12. 11.  1.  4.  6.  7.  9.]
LP_FAMILIE_FEIN
Unique before cleaning:  [ 2. nan  1.  0. 10.  8.  6. 11.  9.  7.  5.  3.  4.]
List: ['0']; Len 1
0
[ 2. nan  1. 10.  8.  6. 11.  9.  7.  5.  3.  4.]
LP_FAMILIE_GROB
Unique before cleaning:  [ 2. nan  1.  0.  5.  4.  3.]
List: ['0']; Len 1
0
[ 2. nan  1.  5.  4.  3.]
LP_STATUS_FEIN
Unique before cleaning:  [10. nan  9.  1.  7.  6.  3.  5.  8.  4.  2.]
List: ['0']; Len 1
0
[10. nan  9.  1.  7.  6.  3.  5.  8.  4.  2.]
LP_STATUS_GROB
Unique before cleaning:  [ 5. nan  4.  1.  3.  2.]
List: ['0']; Len 1
0
[ 5. nan  4.  1.  3.  2.]
NATIONALITAET_KZ
Unique before cleaning:  [1 2 0 3]
List: ['-1', '0']; Len 2
-1
[1 2 0 3]
0
[ 1.  2. nan  3.]
PRAEGENDE_JUGENDJAHRE
Unique before cleaning:  [ 4  0  1  8  9  6 15 14 11  5  3  2 10 13 12  7]
List: ['-1', '0']; Len 2
-1
[ 4  0  1  8  9  6 15 14 11  5  3  2 10 13 12  7]
0
[ 4. nan  1.  8.  9.  6. 15. 14. 11.  5.  3.  2. 10. 13. 12.  7.]
RETOURTYP_BK_S
Unique before cleaning:  [ 5. nan  3.  2.  4.  1.]
List: ['0']; Len 1
0
[ 5. nan  3.  2.  4.  1.]
SEMIO_SOZ
Unique before cleaning:  [6 3 2 4 5 7 1]
List: ['-1', '9']; Len 2
-1
[6 3 2 4 5 7 1]
9
[6 3 2 4 5 7 1]
SEMIO_FAM
Unique before cleaning:  [5 6 2 4 3 1 7]
List: ['-1', '9']; Len 2
-1
[5 6 2 4 3 1 7]
9
[5 6 2 4 3 1 7]
SEMIO_REL
Unique before cleaning:  [2 1 3 4 5 6 7]
List: ['-1', '9']; Len 2
-1
[2 1 3 4 5 6 7]
9
[2 1 3 4 5 6 7]
SEMIO_MAT
Unique before cleaning:  [6 3 4 1 2 5 7]
List: ['-1', '9']; Len 2
-1
[6 3 4 1 2 5 7]
9
[6 3 4 1 2 5 7]
SEMIO_VERT
Unique before cleaning:  [6 7 3 5 2 4 1]
List: ['-1', '9']; Len 2
-1
[6 7 3 5 2 4 1]
9
[6 7 3 5 2 4 1]
SEMIO_LUST
Unique before cleaning:  [7 5 6 4 2 1 3]
List: ['-1', '9']; Len 2
-1
[7 5 6 4 2 1 3]
9
[7 5 6 4 2 1 3]
SEMIO_ERL
Unique before cleaning:  [3 7 4 6 2 5 1]
List: ['-1', '9']; Len 2
-1
[3 7 4 6 2 5 1]
9
[3 7 4 6 2 5 1]
SEMIO_KULT
Unique before cleaning:  [4 1 5 6 7 2 3]
List: ['-1', '9']; Len 2
-1
[4 1 5 6 7 2 3]
9
[4 1 5 6 7 2 3]
SEMIO_RAT
Unique before cleaning:  [1 2 3 5 4 6 7]
List: ['-1', '9']; Len 2
-1
[1 2 3 5 4 6 7]
9
[1 2 3 5 4 6 7]
SEMIO_KRIT
Unique before cleaning:  [3 7 5 1 6 2 4]
List: ['-1', '9']; Len 2
-1
[3 7 5 1 6 2 4]
9
[3 7 5 1 6 2 4]
SEMIO_DOM
Unique before cleaning:  [1 3 5 6 2 4 7]
List: ['-1', '9']; Len 2
-1
[1 3 5 6 2 4 7]
9
[1 3 5 6 2 4 7]
SEMIO_KAEM
Unique before cleaning:  [1 2 6 3 7 4 5]
List: ['-1', '9']; Len 2
-1
[1 2 6 3 7 4 5]
9
[1 2 6 3 7 4 5]
SEMIO_PFLICHT
Unique before cleaning:  [2 4 3 5 1 7 6]
List: ['-1', '9']; Len 2
-1
[2 4 3 5 1 7 6]
9
[2 4 3 5 1 7 6]
SEMIO_TRADV
Unique before cleaning:  [1 4 3 2 5 6 7]
List: ['-1', '9']; Len 2
-1
[1 4 3 2 5 6 7]
9
[1 4 3 2 5 6 7]
SHOPPER_TYP
Unique before cleaning:  [ 3  1  0  2 -1]
List: ['-1']; Len 1
-1
[ 3.  1.  0.  2. nan]
SOHO_KZ
Unique before cleaning:  [ 0. nan  1.]
List: ['-1']; Len 1
-1
[ 0. nan  1.]
TITEL_KZ
Unique before cleaning:  [ 0. nan  1.  4.  3.  5.]
List: ['-1', '0']; Len 2
-1
[ 0. nan  1.  4.  3.  5.]
0
[nan  1.  4.  3.  5.]
VERS_TYP
Unique before cleaning:  [ 1  2 -1]
List: ['-1']; Len 1
-1
[ 1.  2. nan]
ZABEOTYP
Unique before cleaning:  [3 1 2 4 6 5]
List: ['-1', '9']; Len 2
-1
[3 1 2 4 6 5]
9
[3 1 2 4 6 5]
ALTER_HH
Unique before cleaning:  [10. 11.  6.  8. 20.  5. 14. 21. 15. 17.  0. 19.  9. 12. 13. nan 18.  7.
 16.  4.  2.  3.]
List: ['0']; Len 1
0
[10. 11.  6.  8. 20.  5. 14. 21. 15. 17. nan 19.  9. 12. 13. 18.  7. 16.
  4.  2.  3.]
ANZ_PERSONEN
Unique before cleaning:  [ 2.  3.  1.  0.  4.  5.  6. nan  8.  7.  9. 12. 11. 10. 14. 16. 15. 21.
 13.]
ANZ_TITEL
Unique before cleaning:  [ 0. nan  1.  2.  3.  5.]
HH_EINKOMMEN_SCORE
Unique before cleaning:  [ 1. nan  4.  6.  2.  3.  5.]
List: ['-1', '0']; Len 2
-1
[ 1. nan  4.  6.  2.  3.  5.]
0
[ 1. nan  4.  6.  2.  3.  5.]
KK_KUNDENTYP
Unique before cleaning:  [nan  2.  5.  4.  1.  3.  6.]
List: ['-1']; Len 1
-1
[nan  2.  5.  4.  1.  3.  6.]
W_KEIT_KIND_HH
Unique before cleaning:  [ 6.  0. nan  2.  4.  1.  5.  3.]
List: ['-1', '0']; Len 2
-1
[ 6.  0. nan  2.  4.  1.  5.  3.]
0
[ 6. nan  2.  4.  1.  5.  3.]
WOHNDAUER_2008
Unique before cleaning:  [ 9.  3.  7.  8.  5. nan  6.  4.  2.  1.]
List: ['-1', '0']; Len 2
-1
[ 9.  3.  7.  8.  5. nan  6.  4.  2.  1.]
0
[ 9.  3.  7.  8.  5. nan  6.  4.  2.  1.]
ANZ_HAUSHALTE_AKTIV
Unique before cleaning:  [  1.  nan   0.   7.  74.   9.  17.  12.  29.   2.  45.  10.  14.  15.
  55.   6.   5. 108.   3.  25.  18.   4.  11.  19.   8.  26.  43.  63.
  47.  16.  31.  57.  37.  13.  22.  20.  23.  27. 305.  52.  21.  34.
  46.  24.  33.  51. 523. 104. 101.  59.  44.  38.  61.  69.  28.  48.
  40.  39. 145. 107.  32.  76. 117.  87.  84.  41. 286.  92. 222. 119.
  71.  36. 344.  42.  82. 160.  35.  67.  30. 112.  85.  93.  58.  60.
 221.  77. 102. 140.  79.  53. 290. 243. 105.  75.  50.  70.  54. 165.
  86. 156. 277.  95.  68. 185.  66. 253. 157. 155. 158. 149.  72.  99.
 379.  89. 103. 209. 223. 170. 111.  56. 124.  81. 113.  98. 154.  83.
 115. 106.  73. 172.  97.  65. 159.  62.  64. 311. 252. 116. 304. 100.
  94.  78. 229. 110. 246.  49. 181. 147. 137. 163. 321. 114. 138.  80.
 128.  96. 210. 118. 167. 195. 127. 240. 141. 265. 366.  90. 109.  88.
 189. 161. 142. 122. 267. 230.  91. 199. 207. 134. 204. 136. 125. 301.
 307. 200. 120. 139. 171. 331. 179. 133. 146. 258. 202. 135. 274. 121.
 238. 395. 219. 211. 129. 153. 131. 150. 178. 148. 254. 196. 367. 276.
 123. 190. 186. 143. 348. 259. 126.]
List: ['0']; Len 1
0
[  1.  nan   7.  74.   9.  17.  12.  29.   2.  45.  10.  14.  15.  55.
   6.   5. 108.   3.  25.  18.   4.  11.  19.   8.  26.  43.  63.  47.
  16.  31.  57.  37.  13.  22.  20.  23.  27. 305.  52.  21.  34.  46.
  24.  33.  51. 523. 104. 101.  59.  44.  38.  61.  69.  28.  48.  40.
  39. 145. 107.  32.  76. 117.  87.  84.  41. 286.  92. 222. 119.  71.
  36. 344.  42.  82. 160.  35.  67.  30. 112.  85.  93.  58.  60. 221.
  77. 102. 140.  79.  53. 290. 243. 105.  75.  50.  70.  54. 165.  86.
 156. 277.  95.  68. 185.  66. 253. 157. 155. 158. 149.  72.  99. 379.
  89. 103. 209. 223. 170. 111.  56. 124.  81. 113.  98. 154.  83. 115.
 106.  73. 172.  97.  65. 159.  62.  64. 311. 252. 116. 304. 100.  94.
  78. 229. 110. 246.  49. 181. 147. 137. 163. 321. 114. 138.  80. 128.
  96. 210. 118. 167. 195. 127. 240. 141. 265. 366.  90. 109.  88. 189.
 161. 142. 122. 267. 230.  91. 199. 207. 134. 204. 136. 125. 301. 307.
 200. 120. 139. 171. 331. 179. 133. 146. 258. 202. 135. 274. 121. 238.
 395. 219. 211. 129. 153. 131. 150. 178. 148. 254. 196. 367. 276. 123.
 190. 186. 143. 348. 259. 126.]
ANZ_HH_TITEL
Unique before cleaning:  [ 0. nan  2.  4.  1. 13.  6.  5.  3. 20.  9. 11.  8. 14. 18. 23.  7. 12.
 17. 15. 10.]
GEBAEUDETYP
Unique before cleaning:  [ 1. nan  8.  2.  3.  4.  6.]
List: ['-1', '0']; Len 2
-1
[ 1. nan  8.  2.  3.  4.  6.]
0
[ 1. nan  8.  2.  3.  4.  6.]
KONSUMNAEHE
Unique before cleaning:  [ 5.  1.  2.  4.  3.  6. nan  7.]
MIN_GEBAEUDEJAHR
Unique before cleaning:  [1992.   nan 1994. 1997. 1995. 1996. 2011. 2008. 1991. 1990. 2007. 2005.
 1993. 2001. 1999. 2003. 1998. 2000. 2015. 2006. 2004. 1989. 2010. 2002.
 2014. 2012. 2009. 2013. 1987. 1988. 1986. 1985. 2016.]
List: ['0']; Len 1
0
[1992.   nan 1994. 1997. 1995. 1996. 2011. 2008. 1991. 1990. 2007. 2005.
 1993. 2001. 1999. 2003. 1998. 2000. 2015. 2006. 2004. 1989. 2010. 2002.
 2014. 2012. 2009. 2013. 1987. 1988. 1986. 1985. 2016.]
OST_WEST_KZ
Unique before cleaning:  ['W' nan 'O']
List: ['-1']; Len 1
-1
['W' nan 'O']
WOHNLAGE
Unique before cleaning:  [ 7. nan  2.  3.  1.  4.  5.  8.  0.]
List: ['-1']; Len 1
-1
[ 7. nan  2.  3.  1.  4.  5.  8.  0.]
CAMEO_DEUG_2015
Unique before cleaning:  ['1' nan '5' '4' '7' '3' '9' '2' '6' '8' 'X']
List: ['-1', 'X']; Len 2
-1
['1' nan '5' '4' '7' '3' '9' '2' '6' '8' 'X']
X
['1' nan '5' '4' '7' '3' '9' '2' '6' '8']
CAMEO_DEU_2015
Unique before cleaning:  ['1A' nan '5D' '4C' '7B' '3B' '1D' '9E' '2D' '4A' '6B' '9D' '8B' '5C' '9C'
 '4E' '6C' '8C' '8A' '5B' '9B' '3D' '2A' '3C' '5F' '7A' '1E' '2C' '7C'
 '5A' '2B' '6D' '7E' '5E' '6E' '3A' '9A' '4B' '1C' '1B' '6A' '8D' '7D'
 '6F' '4D' 'XX']
List: ['XX']; Len 1
XX
['1A' nan '5D' '4C' '7B' '3B' '1D' '9E' '2D' '4A' '6B' '9D' '8B' '5C' '9C'
 '4E' '6C' '8C' '8A' '5B' '9B' '3D' '2A' '3C' '5F' '7A' '1E' '2C' '7C'
 '5A' '2B' '6D' '7E' '5E' '6E' '3A' '9A' '4B' '1C' '1B' '6A' '8D' '7D'
 '6F' '4D']
CAMEO_INTL_2015
Unique before cleaning:  ['13' nan '34' '24' '41' '23' '15' '55' '14' '22' '43' '51' '33' '25' '44'
 '54' '32' '12' '35' '31' '45' '52' 'XX']
List: ['-1', 'XX']; Len 2
-1
['13' nan '34' '24' '41' '23' '15' '55' '14' '22' '43' '51' '33' '25' '44'
 '54' '32' '12' '35' '31' '45' '52' 'XX']
XX
['13' nan '34' '24' '41' '23' '15' '55' '14' '22' '43' '51' '33' '25' '44'
 '54' '32' '12' '35' '31' '45' '52']
KBA05_ANTG1
Unique before cleaning:  [ 2. nan  3.  0.  4.  1.]
List: ['-1']; Len 1
-1
[ 2. nan  3.  0.  4.  1.]
KBA05_ANTG2
Unique before cleaning:  [ 2. nan  0.  3.  1.  4.]
List: ['-1']; Len 1
-1
[ 2. nan  0.  3.  1.  4.]
KBA05_ANTG3
Unique before cleaning:  [ 0. nan  2.  1.  3.]
List: ['-1']; Len 1
-1
[ 0. nan  2.  1.  3.]
KBA05_ANTG4
Unique before cleaning:  [ 0. nan  2.  1.]
List: ['-1']; Len 1
-1
[ 0. nan  2.  1.]
KBA05_BAUMAX
Unique before cleaning:  [ 0. nan  1.  5.  3.  4.  2.]
List: ['-1', '0']; Len 2
-1
[ 0. nan  1.  5.  3.  4.  2.]
0
[nan  1.  5.  3.  4.  2.]
KBA05_GBZ
Unique before cleaning:  [ 4. nan  3.  1.  5.  2.  0.]
List: ['-1', '0']; Len 2
-1
[ 4. nan  3.  1.  5.  2.  0.]
0
[ 4. nan  3.  1.  5.  2.]
BALLRAUM
Unique before cleaning:  [ 3. nan  7.  6.  5.  2.  1.  4.]
List: ['-1']; Len 1
-1
[ 3. nan  7.  6.  5.  2.  1.  4.]
EWDICHTE
Unique before cleaning:  [ 2. nan  4.  1.  5.  3.  6.]
List: ['-1']; Len 1
-1
[ 2. nan  4.  1.  5.  3.  6.]
INNENSTADT
Unique before cleaning:  [ 4. nan  1.  7.  8.  5.  6.  2.  3.]
List: ['-1']; Len 1
-1
[ 4. nan  1.  7.  8.  5.  6.  2.  3.]
GEBAEUDETYP_RASTER
Unique before cleaning:  [ 4. nan  3.  2.  5.  1.]
KKK
Unique before cleaning:  [ 1. nan  3.  4.  2.  0.]
List: ['-1', '0']; Len 2
-1
[ 1. nan  3.  4.  2.  0.]
0
[ 1. nan  3.  4.  2.]
MOBI_REGIO
Unique before cleaning:  [ 4. nan  3.  1.  5.  2.  6.]
ONLINE_AFFINITAET
Unique before cleaning:  [ 3. nan  1.  2.  5.  4.  0.]
REGIOTYP
Unique before cleaning:  [ 1. nan  7.  6.  3.  4.  2.  5.  0.]
List: ['-1', '0']; Len 2
-1
[ 1. nan  7.  6.  3.  4.  2.  5.  0.]
0
[ 1. nan  7.  6.  3.  4.  2.  5.]
KBA13_ANZAHL_PKW
Unique before cleaning:  [1201.   nan  433. ...   77.   13.   34.]
PLZ8_ANTG1
Unique before cleaning:  [ 3. nan  2.  1.  4.  0.]
List: ['-1']; Len 1
-1
[ 3. nan  2.  1.  4.  0.]
PLZ8_ANTG2
Unique before cleaning:  [ 3. nan  2.  4.  1.  0.]
List: ['-1']; Len 1
-1
[ 3. nan  2.  4.  1.  0.]
PLZ8_ANTG3
Unique before cleaning:  [ 1. nan  3.  2.  0.]
List: ['-1']; Len 1
-1
[ 1. nan  3.  2.  0.]
PLZ8_ANTG4
Unique before cleaning:  [ 0. nan  1.  2.]
List: ['-1']; Len 1
-1
[ 0. nan  1.  2.]
PLZ8_BAUMAX
Unique before cleaning:  [ 1. nan  3.  2.  5.  4.]
List: ['-1', '0']; Len 2
-1
[ 1. nan  3.  2.  5.  4.]
0
[ 1. nan  3.  2.  5.  4.]
PLZ8_HHZ
Unique before cleaning:  [ 5. nan  3.  4.  2.  1.]
List: ['-1']; Len 1
-1
[ 5. nan  3.  4.  2.  1.]
PLZ8_GBZ
Unique before cleaning:  [ 5. nan  2.  4.  3.  1.]
List: ['-1']; Len 1
-1
[ 5. nan  2.  4.  3.  1.]
ARBEIT
Unique before cleaning:  [ 1. nan  3.  2.  4.  5.  9.]
List: ['-1', '9']; Len 2
-1
[ 1. nan  3.  2.  4.  5.  9.]
9
[ 1. nan  3.  2.  4.  5.]
ORTSGR_KLS9
Unique before cleaning:  [ 2. nan  5.  3.  7.  4.  8.  6.  1.  9.]
List: ['-1', '0']; Len 2
-1
[ 2. nan  5.  3.  7.  4.  8.  6.  1.  9.]
0
[ 2. nan  5.  3.  7.  4.  8.  6.  1.  9.]
RELAT_AB
Unique before cleaning:  [ 1. nan  3.  5.  2.  4.  9.]
List: ['-1', '9']; Len 2
-1
[ 1. nan  3.  5.  2.  4.  9.]
9
[ 1. nan  3.  5.  2.  4.]
Done.
2/3  Removing selected columns and rows...
Done.
3/3  Selecting, re-encoding, and engineering column values...
Done.
Data cleaned successfully.
========================================== /Data cleaning ================================================



(136159, 194)
(774743, 194)
Missing values
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.



Scaling
PCA
K-means

#print(set(customers_prediction))

#print(df_customers_pca['GEBAEUDETYP_5.0'])

print(df_customers_pca)

[[-6.80046859  2.68587516  4.11647363 ... -0.88962619 -0.58691065
   0.35775874]
 [-1.6261704   5.6610524  -0.65003148 ...  0.18034395  0.25105177
  -0.23562118]
 [-5.55963543  0.91505592  2.23342051 ...  1.38759941 -0.80673676
  -0.22132898]
 ...
 [-5.61904618  3.51341768  2.69156509 ...  0.19935001  0.21444607
  -0.74407117]
 [-1.56096952  2.60919726 -2.49763285 ...  1.2616106  -0.48860208
  -0.5687784 ]
 [-5.14908607 -2.1435113   2.24471119 ...  0.8213381  -0.58168532
   0.19190296]]

Step 3.3: Compare Customer Data to Demographics Data

At this point, you have clustered data based on demographics of the general population of Germany, and seen how the customer data for a mail-order sales company maps onto those demographic clusters. In this final substep, you will compare the two cluster distributions to see where the strongest customer base for the company is.

Consider the proportion of persons in each cluster for the general population, and the proportions for the customers. If we think the company's customer base to be universal, then the cluster assignment proportions should be fairly similar between the two. If there are only particular segments of the population that are interested in the company's products, then we should see a mismatch from one to the other. If there is a higher proportion of persons in a cluster for the customer data compared to the general population (e.g. 5% of persons are assigned to a cluster for the general population, but 15% of the customer data is closest to that cluster's centroid) then that suggests the people in that cluster to be a target audience for the company. On the other hand, the proportion of the data in a cluster being larger in the general population than the customer data (e.g. only 2% of customers closest to a population centroid that captures 6% of the data) suggests that group of persons to be outside of the target demographics.

Take a look at the following points in this step:

• Compute the proportion of data points in each cluster for the general population and the customer data. Visualizations will be useful here: both for the individual dataset proportions, but also to visualize the ratios in cluster representation between groups. Seaborn's countplot() or barplot() function could be handy.
  • Recall the analysis you performed in step 1.1.3 of the project, where you separated out certain data points from the dataset if they had more than a specified threshold of missing values. If you found that this group was qualitatively different from the main bulk of the data, you should treat this as an additional data cluster in this analysis. Make sure that you account for the number of data points in this subset, for both the general population and customer datasets, when making your computations!
• Which cluster or clusters are overrepresented in the customer dataset compared to the general population? Select at least one such cluster and infer what kind of people might be represented by that cluster. Use the principal component interpretations from step 2.3 or look at additional components to help you make this inference. Alternatively, you can use the .inverse_transform() method of the PCA and StandardScaler objects to transform centroids back to the original data space and interpret the retrieved values directly.
• Perform a similar investigation for the underrepresented clusters. Which cluster or clusters are underrepresented in the customer dataset compared to the general population, and what kinds of people are typified by these clusters?

# Compare the proportion of data in each cluster for the customer data to the
# proportion of data in each cluster for the general population.

fig, axs = plt.subplots(nrows=1, ncols=2, figsize=(15, 5))
sns.countplot(x=population_prediction, label="Population", ax=axs[0])
axs[0].set_title("Population clusters histogram")
sns.countplot(x=customers_prediction, label="Customers", ax=axs[1])
axs[1].set_title("Customers clusters histogram")

Text(0.5, 1.0, 'Customers clusters histogram')

# What kinds of people are part of a cluster that is overrepresented in the
# customer data compared to the general population?

def describe_cluster(k):
    '''
    k is the number of cluster counting from 0 (like ticks at the plots above)
    '''
    
    print("Cluster {} (tick {})".format(k+1, k))
    
    # Checking for the most correlated  and most inversely correlated columns
    w = pca_weights_mapping(df_customers_std, pca125, k)
    #print(w, "\n")
    
    #inverse transformation
    c = scaler.inverse_transform(pca125.inverse_transform(df_customers_pca[np.where(customers_prediction==k)])).round()
    c_df = pd.DataFrame(data=c, index=np.array(range(0, c.shape[0])), columns=df_customers.columns)

    #getting info
    indexes= list(w.index[0:3])
    indexes.extend(w.index[-3:])
    #print(indexes)

    cols_for_describe_df = ['Mean', 'Median', 'Mode']
    
    dict_describe = {'Mean': [], 'Median': [], 'Mode': []}
    
    for col in indexes:
        dict_describe['Mean'].append(c_df[col].mean())
        dict_describe['Median'].append(c_df[col].median())
        dict_describe['Mode'].append(c_df[col].mode()[0])
        
    #print(dict_describe)
    
    describe = pd.DataFrame(data=dict_describe, index=indexes, columns=cols_for_describe_df)
    print(describe)
    print("\n\n")
    
    
# Cluster 4 (tick 3)
describe_cluster(3)


# Cluster 9 (tick 8)
describe_cluster(8)



# Cluster 12 (tick 11)
describe_cluster(11)

Cluster 4 (tick 3)
                        Mean  Median  Mode
GREEN_AVANTGARDE    0.610246     1.0   1.0
PJ_MOVEMENT         0.610246     1.0   1.0
LP_STATUS_GROB_5.0  0.152226     0.0   0.0
LP_STATUS_FEIN_9.0  0.017708    -0.0  -0.0
LP_STATUS_GROB_4.0  0.019845     0.0  -0.0
WOHNLAGE            3.435611     3.0   3.0



Cluster 9 (tick 8)
                        Mean  Median   Mode
OST_WEST_KZ         0.089968     0.0    0.0
CI_LIFESTAGE        3.249210     3.0    4.0
ARBEIT              2.430296     2.0    2.0
PLZ8_GBZ            3.916643     4.0    4.0
PLZ8_HHZ            3.555045     3.0    3.0
KBA13_ANZAHL_PKW  717.823513   656.0  568.0



Cluster 12 (tick 11)
                         Mean  Median  Mode
CI_LIFESTAGE         1.875933     1.0   1.0
CAMEO_DEUG_2015_6    0.078358     0.0  -0.0
CAMEO_DEUG_2015_3    0.006530     0.0  -0.0
LP_FAMILIE_FEIN_2.0  0.101679    -0.0  -0.0
CAMEO_DEU_2015_7A    0.045709     0.0   0.0
CAMEO_DEUG_2015_7    0.148321     0.0  -0.0

# What kinds of people are part of a cluster that is underrepresented in the
# customer data compared to the general population?


# Cluster 11 (tick 10)
describe_cluster(10)


# Cluster 7 (tick 6)
describe_cluster(6)


# Cluster 8 (tick 7)
describe_cluster(7)

Cluster 11 (tick 10)
                         Mean  Median  Mode
LP_FAMILIE_GROB_2.0  0.193856     0.0  -0.0
LP_FAMILIE_FEIN_2.0  0.193856     0.0  -0.0
LP_STATUS_GROB_3.0   0.003199     0.0  -0.0
KBA05_ANTG4          0.005163    -0.0   0.0
LP_STATUS_GROB_2.0   0.004763     0.0   0.0
LP_FAMILIE_GROB_4.0  0.057917     0.0  -0.0



Cluster 7 (tick 6)
                        Mean  Median  Mode
GEBAEUDETYP_3.0     0.077402     0.0  -0.0
LP_STATUS_GROB_3.0  0.095519     0.0   0.0
LP_STATUS_FEIN_6.0  0.017544    -0.0   0.0
CAMEO_DEUG_2015_6   0.115181     0.0  -0.0
GEBAEUDETYP_RASTER  4.000822     4.0   4.0
GEBAEUDETYP_1.0     0.777587     1.0   1.0



Cluster 8 (tick 7)
                          Mean  Median  Mode
LP_FAMILIE_GROB_4.0   0.116570     0.0   0.0
LP_STATUS_FEIN_3.0    0.283930     0.0  -0.0
LP_FAMILIE_FEIN_7.0   0.043297    -0.0   0.0
LP_FAMILIE_FEIN_10.0  0.138218     0.0   0.0
CAMEO_DEUG_2015_9     0.160699     0.0  -0.0
LP_FAMILIE_GROB_5.0   0.220649     0.0   0.0

Discussion 3.3: Compare Customer Data to Demographics Data

The above diagrams show that the top three bars for Population and the top three bars for Customers correspond to the same clusters: clusters no. 4, 9 and 12 (in ticks: 3, 8, 11). Thus, the company's 3 most important target groups coincide with the three most numerous groups in the population. A large target audience is a greater likelihood of high income for the company. This distribution of customer groups is promising. However, there is room for improvement. Namely, the group with the lowest share among customers is claster 11 (tick 10 on the charts). For the population, however, cluster 11 is consecutively the fourth most numerous. Also, clusters 7 and 8 are under-represented among customers.

About question 1: "What kinds of people are part of a clusters that are overrepresented in the customer data compared to the general population?"

Cluster 4 (tick 3)

FEATURE	Mean	Median	Mode	What the feature describes	What the value tells (characteristic of people in cluster)
PJ_MOVEMENT	0.974209	1.0	1.0	movement of person's youth	avant-garde
GREEN_AVANTGARDE	0.974209	1.0	1.0	if membership in environmental sustainability as part of youth	yes, green avant-garde member
LP_STATUS_GROB_5.0	0.850048	1.0	1.0	if social status = top earners	yes
LP_STATUS_FEIN_9.0	0.001850	0.0	0.0	if social status = houseowner	no
LP_STATUS_GROB_4.0	0.002229	0.0	0.0	if social status = houseowner (double feature)	no
WOHNLAGE	3.054167	3.0	3.0	neighborhood quality	average neighborhood

Generally, cluster 4 is mainly related to people growing up in the time of avant-garde, involved in the green movement, with very high incomes, living in an average neighborhood.

Cluster 9 (tick 8)

FEATURE	Mean	Median	Mode	What the feature describes	What the value tells (characteristic of people in cluster)
OST_WEST_KZ	0.067668	-0.0	-0.0	Building location via former East / West Germany (GDR / FRG)	West (FRG)
CI_LIFESTAGE	3.270293	3.0	4.0	life stage typology	Older Families & Mature Couples
ARBEIT	2.242643	2.0	2.0	Share of unemployment in community	low
PLZ8_GBZ	3.881307	4.0	4.0	Number of buildings within the region	300-449 buildings
PLZ8_HHZ	3.382820	3.0	3.0	Number of households within the PLZ8 region	300-599 households
KBA13_ANZAHL_PKW	699.099667	638.0	536.0	Number of cars in the region	around 640

Cluster 9 is mainly associated with people living in West Germany (former FRG) belonging to the group of older families or mature couples, living in a community of low unemployment rate, rather in a village or a small town (based on number of houses, number of households and number of cars).

Cluster 12 (tick 11)

FEATURE	Mean	Median	Mode	What the feature describes	What the value tells (characteristic of people in cluster)
CI_LIFESTAGE	2.836338	3.0	1.0	life stage typology	Pre-Family Couples & Singles
CAMEO_DEUG_2015_6	0.228653	-0.0	0.0	if low-consumption middleclass	no
CAMEO_DEUG_2015_3	0.054887	0.0	-0.0	if established middleclass	no
LP_FAMILIE_FEIN_2.0	0.262084	0.0	-0.0	if couple (as status of relationship)	no
CAMEO_DEU_2015_7A	0.054762	0.0	0.0	if journeymen	no
CAMEO_DEUG_2015_7	0.166220	-0.0	-0.0	if lower middleclass	no

Cluster 12 is mainly associated with people who are single, not low-consumption middleclass or established middleclass, not journeymen and not belonging to the lower middleclass.

About question 2: "What kinds of people are part of a clusters that are underrepresented in the customer data compared to the general population?"

Cluster 11 (tick 10)

FEATURE	Mean	Median	Mode	What the feature describes	What the value tells (characteristic of people in cluster)
LP_FAMILIE_GROB_2.0	0.104953	0.0	0.0	if family type = couple	no
LP_FAMILIE_FEIN_2.0	0.104953	0.0	0.0	if couple (as status of relationship)	no
LP_STATUS_GROB_3.0	0.028302	0.0	-0.0	if social status = independents	no
KBA05_ANTG4	0.706368	0.0	0.0	Number of 10+ family houses in the microcell	no any
LP_STATUS_GROB_2.0	0.053066	0.0	0.0	if social status = average earners	no
LP_FAMILIE_GROB_4.0	0.083726	0.0	-0.0	if family type = single family	no

The selected most correlated features in this case say who the people in cluster 11 ARE NOT. People are not in an informal relationship (but this does not mean that they are not in any relationship at all), they are not average earners (they can earn less or more), are not a single family in a household and do not have multi-family buildings in their neighborhood (+10 families).

Cluster 7 (tick 6)

FEATURE	Mean	Median	Mode	What the feature describes	What the value tells (characteristic of people in cluster)
GEBAEUDETYP_3.0	0.423174	-0.0	-0.0	if type of building = mixed (=residential and company) building	no
LP_STATUS_GROB_3.0	0.053736	0.0	0.0	if social status = independents	no
LP_STATUS_FEIN_6.0	0.021830	0.0	0.0	if social status = independent workers	no
CAMEO_DEUG_2015_6	0.109992	0.0	-0.0	if low-consumption middleclass	no
GEBAEUDETYP_RASTER	3.109152	3.0	3.0	Ratio of residential to commercial activity	mixed cell with middle business share
GEBAEUDETYP_1.0	0.314861	-0.0	0.0	if type of building = residential building	no

The selected most correlated features in this case say who the people in cluster 11 ARE NOT. People are not living in mixed buildings (residental and company), are not independent workers, are not low-consumption middleclass, and ARE living in the area of middle business share.

Cluster 8 (tick 7)

FEATURE	Mean	Median	Mode	What the feature describes	What the value tells (characteristic of people in cluster)
LP_FAMILIE_GROB_4.0	0.069121	-0.0	-0.0	if family type = single family	no
LP_STATUS_FEIN_3.0	0.060105	-0.0	0.0	if social status = aspiring low-income earners	no
LP_FAMILIE_FEIN_7.0	0.027799	-0.0	-0.0	if family with teenager	no
LP_FAMILIE_FEIN_10.0	0.190083	0.0	-0.0	if two-generational household	no
CAMEO_DEUG_2015_9	0.018032	0.0	0.0	if urban working class	no
LP_FAMILIE_GROB_5.0	0.301277	0.0	0.0	if family type = multiperson household	no

The selected most correlated features in this case say who the people in cluster 11 ARE NOT. People are not living as a single full family, are not aspiring low-income earners, are not family with teenager, are not living in a two-generational household, are not urban working class and are not living in a multiperson household.

---

Underrepresented clusters's most corelated features do not make it clear to which groups of people they apply. The specification based on them is not clear and intuitive. A much more intuitive grouping can be seen in the case of overrepresented clusters, which described above.

Congratulations on making it this far in the project! Before you finish, make sure to check through the entire notebook from top to bottom to make sure that your analysis follows a logical flow and all of your findings are documented in Discussion cells. Once you've checked over all of your work, you should export the notebook as an HTML document to submit for evaluation. You can do this from the menu, navigating to File -> Download as -> HTML (.html). You will submit both that document and this notebook for your project submission.